ТНЕ TRANSACTIONS OF THE LINNEAN SOCIETY OF LON DON. d > E E EI < SECOND SERIES—VOLUME BOTANY. LONDON: а FRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE ; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER- ROW Аосозт 1880. AGE T Nine EY 2. MISSOURI nuni Р уто CONTENTS. PART L- Jory, 1875. I. On Napoleona, Omphalocarpum, and Asteranthos. Ву .онх MIERS, F.R.S., V.P.L.S., Dignit. et Commend. Ord. Imp. Bras. Rose, $c. (Plates I.-IV.) page 1 II. On the Auxemme:x, а new Tribe of the Cordiacee. Ву Тонх Miers, F.R.S., Р.В. де. (Plita V VILE) ж, з а S ec. 0 PART II.—DECEMBAER, 1875. III. On the Origin of the prevailing Systems of Phyllotaxis. By the Rev. GEORGE Hu MILES Е). ... ....... 27 IV. On the Barringtoniacee. Ву Јонх Мтевз, F.R.S., F.L.S., Diguit. et Commend. Ord. Imp. Bras. Rose, Fc. (Plates X.-XVIIL) . 2...5... 4 PART III.—JvuswE, 1876. V. On а Collection of Fungi made by Mr. 8сірі? Kurz, Curator of the Botanic Garden, Calcutta. By FREDERICK Currey, M.A., F.R.S., Sec.L.S. (Plates wie Uu. . a о. H9 VI. Preliminary Note on the Rate of Growth of the Female Flower-stalk of Vallisneria spiralis, Linn. By AurreD W. Bennett, M.A., B.Sc., F.L.S., Lecturer on Botany at St. Thomas’s Hospital. (With a Жы... 1685 L’ i УП. On the Growth of the Flower-stalk of the Hyacinth. By ALFRED W. BENNETT, M.A., B.Sc., F.L.S., Lecturer on Botany at St. Thomas's Hospital. (With two иие ои VIII. New British Lichens. Ву the Rev. W.A. LEIGHTON, B.A. Camb., F.L.S., F.B.S. а IX. On the Hygroscopic Mechanism by which certain Seeds are enabled to bury them- selves in the Ground. By FRANCIS Darwin, М.В., F.L.S. (Plate XXIIL, and ENDE ass S 4 oS 20 s 0. s TB PART ТУ. — ресеМВЕК, 1876. X. On the African Species of the Genus Coffea, Linn. Ay W. P. Нтикм, МХ. РЕ Eie ANI a. c QNI C c... .. . 138 ХІ. On the Origin of Floral ZEstivations. With Notes on the Structure of the Crucife- | rous Flower, on that of Adoxa, and on the Corolla of Primula. By the Rev. Но, ҒА Ға, XO. (Plate XXV) 1. IT XII. Оз» some Points in the Histology of certain Species of Corallinacee. By Major- General В. J. Newson, R.E., and Prof. Duncan (М.В. Lond.), F. R.S., Pres. Geol. Soc. (Communicated by Dr. Murre, F.L.S.) (Plates XXVI. & АУ. 107 ХПІ. Ол the Minute Structure and Mode of Growth of Ballia callitricha, Ag. (sensu latiori). Бу ҮҮ. Авснев, F.R.S., M.R.I.A. (Communicated by W.T. THISELTON DxzR, M4; F.L.S.) (Plates XXVIII. & XXIX., and а Woodeut) . . 911 PART V.—JANUARY, 1878. XIV. Оп the Geographical Distribution of the Meliaceæ. By M. CASIMIR DE CAN- DOLLE, of Geneva. (Communicated by G. BENTHAM, Р.Е.8., V.P.L.S.) (Table and Map=Plates ХХХ. & ХХХІ). rye 233 XV. New British Lichens. By the Rev. W. A. LEIGHTON, B.A. Camb., F.L.S., F_B.S. Edinb., фе. (Piste XAS. 2-04 62... . , 987 XVI. New Irish Lichens. By the Rev. W. A. LzrGHTON, B.A. Camb., F.L.S., F.B.S. ‚ Mand. шал 4.4 0.4 XVII. Report on the Liliacee, Iridacee, Hypoxidaceea, and Hemodoracee of Wel- witsel's Angolan Herbarium. | By J. а. Bakxg, F.L.S. (Plates XXXIV.- RASV) ааа E SUA ос. WE [v] XVIII. Contribution to the Lichenographia of New Zealand. By CHARLES KNIGHT, F.L.S., Auditor-General of New Zealand. (Plates XXXVII. & XXXVIII.) page 275 XIX. On some Points in the Morphology of the Primulacee. Ву М.Т. Masters, M.D.» F.R.S. F.L.S., yc. (Plates XXXIX.-XLL., and а Woodcut.). . . . . 285 PART VI.—Janvaky, 1879. XX. On Mycoidea parasitica, а new Genus of Parasitic Alge, and the Part which и plays in the Formation of certain Lichens. By Davin DOUGLAS CUNNINGHAM, M.B., F.L.S., Surgeon H.M. Indian Army. (Plates XLII. & XLIII.). . 301 ХХІ. On the Self-fertilization of Plants. By the Rev. Georcr Henstow, M.A., F.L.S., BENE ее EE XXII. List of Fungi from Brisbane, Queensland ; with Descriptions of New Species. Бу the Rev. M. J. BERKELEY, M.A., F. L.S., and С. E. Broome, Р.Г... (Plates XLV. еви и ео M XXIII. On the Occurrence of Conidial Fructification in the Mucorini, illustrated by Choanephora. Ву D. D. Cunninenam, M.B., F.L.S., Surgeon Н.М. Indian кені серме ХҮЛІНБУ ee a ee 2 т с MIS PART УП.— ЕЕвволву, 1880. XXIV. On а new Species of Helvella. Ву Упллам Рнпллрѕ, F.L.S. (Plate 212942502 5-0... WM XXV. A Review of the Ferns of Northern India. (Part I.) By CHARLES BARON CLABEE, MA. FDS: CU Зы. occ 4% PART VIII.—APnir, 1880. XXV. А Review of the Ferns of Northern India. (Part П.) Ву CHARLES BARON Dus MAE Г. ПА ГАН)... о... 486 PART IX.—Aveust, 1880. | XXV. A Review of the Ferns of Northern India. (Part ПІ.) Бу CHARLES BARON CLARKE, MA, F.L.S. (Plates LXXIV.—LXXXIV.). . . . . . page 567 XXVI. On the Origin of the so-called Scorpioid Cyme. Ву the Rev. GEORGE HgNsrow, Е EGGS PCPS. БАХ) ....... UB INDEX AND Contents то Vor. Г. ERRATA. Page 94, line 3 from bottom, for * Pirigara” read Perigara. 115, line 4 PE * Butonica rotata ” read Butonica rosata. 124, line 21 from top, for “ А. CINERASCENS " read Р. CINERASCENS. 287, by inadvertence of printer, is made ** 728." 444, line 18 from top, omit * Cystopteris dimidiata, Dene. іп Jacquem. Voy. Bot. t. 178,” which is rightly referred to DAVALLIA BULLATA in p. 445. 446, line 13 from bottom, in “ serratures " a letter has accidentally dropped out. 504, top line, ** Hemedictyum” should be Hemidictyum. | 512, bottom line, for ** Е. созртолтом ” read N. созртрАТОМ. q 568, line 18 from top, for “ C. Totta” read G. Totta. | ТНЕ TRANSACTIONS THE LINNEAN SOCIETY OF LONDON. SECOND SERIES.—BOTANY. VOLUME I. ` PART THE FIRST. LONDON: PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. M.DCCC.LXXV. CONTENTS. PART 1.—1875. I. Ой Napoleona, Omphalocarpum, and Asteranthos. By Joun Miers, Esq., F.R.S., V.P.L.S., Dignit. et Commend. Ord. Imp. Bras. Rose, ўс. . . . . . pagel IL. On the Auxemme:e, и new Tribe of the Cordiacere. Ву Јонх Miers, Esq., F.R.S., ЛТ 22... 7Ж% TRANSACTIONS OF THE LINNEAN SOCIETY. I. Оп Napoleona, Omphalocarpum, and Asteranthos. By Тонх p БР di, Dignit. et Commend. Ord. Imp. Bras. Rose, фе. (Plates I.-IV.) Read June 4th, 1874. ‘THE plants arranged in the small group of the Napoleonee consist of two heterogeneous genera, the position of which in the system has long been a puzzle to botanists. One of them is of African, the other of Brazilian origin. Тһе former, Napoleona, was established by Palisot de Beauvais for a beautiful species discovered by him in 1787, at which time his drawing and description were made, though not published till 1804'. Asteranthos, at first thought to have been brought from Africa, was afterwards found to be of South- American growth. Desfontaines, who first described and figured the latter in 18907, considered that the two genera should constitute a group belonging to Symplocinee. Desvaux changed the name Napoleona into Belvisia, in honour of Palisot de Beauvais °. Desfontaines subsequently, as well as Brongniart, placed this group between Memecylacee and Rhizophoracee. Brown, in 18224, proposed it as a distinct family (Belvisiee), to stand near Symplocacee, with a tendency towards Rafflesiacee. Lindley in 18335, and again in 1836°, placed Belvisiacee after Campanulacee and Sphenocleniacee, and before Columelliacee ; but in 18457 he changed this view, arranging it between Rhizophoracee and Melastomacee. De Candolle, іп 1839*, located the Napoleonee after Sphenoclenec and Columelliacee, and before Гассииее. Endlicher, in 1839 °, regarded the Napoleonee as doubtful genera of the Hbenacee ; he arranged them provisionally between Symplocee and Columelliacee. Meissner, in 1843 ", placed the Belvisiee between Passifloree and Loasacee. Jussieu, in 1844", in describing a second species of Napoleona, placed this : Fl. Owar. ii. 32, tab. 68. * Mém. Mus. vi. 9, tab. 3. * Journ. de Bot. appl. iv. 130. * Linn. Trans. xiii. 222. 5 Nixus and Introd. Bot. р. 230. * Nat. Syst. p. 239. 7 Veg. King. р. 728. * Prodr. vii. 550. * Gen. Pl. р. 745. ? Gen. PL п Ann, Бе. Nat, 3™ sér. ii. 227, tab. 4. SECOND SERIES.— BOTANY, VOL. I. B 2 МВ. 7. MIERS ON NAPOLEONA, . group between Cucurbitacee and Passifloree. Lindley, in 1845, described and figured a third species of Napoleona, collected by Whitfield in Sierra Leone'. Sir Wm. Hooker and Dr. Planchon, in 1848”, first suggested that the Napoleonee approached nearer to Myrtacee, as shown by the relation of Napoleona to Gustavia in Lecythidacee, and to Luffa in Cucurbitacee ; but they offered no evidence in support of so novel a view. Mr. Bentham’, іп 1849, adopted the same opinion, making the Napoleonee a suborder of Myrtacee, allied to Barringtoniee, on account of its numerous stamens seated upon a disk over an inferior ovary. Agardh, in 1859*, attempted to show that the Napoleonee are allied to Ternstremiacee. Messrs. Hooker and Bentham in 1867”, maintaining their former views, arranged the Napoleonee аз a third subtribe of their Lecythidee, which is their fourth tribe of their IWyrtacee, thus associating it intimately with Barringtoniee and Lecythidee proper. Finally, Dr. Masters, in 18685, contributed a valuable memoir on Napoleona, drawn up from his examination of a flowering species growing at Kew. Its structure is there minutely described ; but he offered no opinion of the true affinity of the group, except in accepting the view of Bentham and Hooker, and in concluding that the Vyrtacee is its more fitting resting-place. We may attribute this extraordinary divergence in the opinions of botanists entirely to the confused notions entertained about the several floral parts, which had been regarded as modified petals by some, as a kind of compound paracorolla by many, while others have considered these parts transformed stamens, or nectariferous emanations from a disk, or coronal appendages analogous to those of Passiflora. А very careful examination of these two genera has convinced me that all have been wide of the mark in regard to their true affinity. In order, therefore, to arrive at a proper conclusion, it is desirable to scrutinize the floral and carpological structure with greater circumspection than has hitherto been employed. 1. NAPOLEONA. In conducting this analysis, it appeared to me that the first essential condition was to separate the different parts from one another. With this view, after moistening a dried flower, a knife was carefully introduced beneath the corolla, so as to cut off its connexion with the disk, which supports all the floral parts; the sepals and ovary were thus set free. The corolla was then carefully detached from the corona in the same manner. This, in the bud state, is plicated and incurved, so as to form a depressed globe ; but when fully developed, and thus separated, it is seen to be gamopetalous, petaloid, orbicular, quite rotate, marked beneath by several subulate flat subprominent radiating nerves, con- nivent at the base, varying in the several species from 30 to 40 (the intermediate spaces being membranaceous) ; and they terminate in as many short obtuse peripherial segments of the border. і The corona, now isolated, is seen to consist of three distinct whorls, all affixed, like the corolla, upon the outer surface of the disk. "The outer whorl, now fully exposed to view, as seen from below, consists of 60 to 70 distinct, narrow, pointed segments, free to the * Bot. Reg. (1844) p. 780; Gard. Chron. 1844, р. 480; Veg. Kingd. (cum icone) p. 728. * Hook. Icon. vii. tab. 799. 3? Niger Flora, p. 361, tab. 49. * Theor. p. 132. * Gen. Plant. i. 696. 5 Journ. Linn. Soc. x. 492, OMPHALOCARPUM, AND ASTERANTHOS. : 3 base, somewhat shorter than the corolla, all radiating outwards, each with a fine median nerve. Ina circle, about 1 of an inch beyond the points of their attachment, we now observe a very distinct moniliform ring of prominent vesicles, one upon each segment, bearing the appearance of as many sterile anther-cells. This important feature appears to me to afford a key to the nature of the whole structure, and seems to have been over- looked by all botanists, except perhaps by Dr. Masters’. The second whorl of the corona is also petaloid, formed of 40 to 60 similar but broader segments, sometimes as long as, often shorter than the corolla. These are confluent along their margins towards the base for $ or $ of their length, thus forming a cup; but the external portions are quite free, erect, or curving inwards. Looking at this cup from below, we see a similar ring of prominences as in the former case, and at the same distance from the base, all forming as many distinct hollow patches, like corrugated vesicles. The third or internal whorl consists of a series of linear segments like the preceding, and, I believe, constantly 20 in number. They are attached to the disk in a similar manner, are quite free to their base, their margins touching or overlapping each other ; they have a fine nerve along their centre up to a short distance below the apex, there terminating in а prominent point, where the anther is attached; these have been con- sidered true filaments by all botanists. Тһе anther is oblong, a little curved, appear- ing 2-lobed, owing to a deep furrow down their middle; it opens bivalvately at the bottom of the furrow, its margins becoming detached from a prominent line upon the connective; before dehiscence, they are consequently 2-locellate; after bursting they appear 1-locular; this cell is adnate for its whole length, upon a narrow fleshy con- nective and is affixed by one of its extremities to the point in the nerve of the filament before mentioned, always extrorsely. By the hippocrepiform curvature of the filaments the anthers become lodged in the 5 open cavities within the area of the concave disk, formed by the projecting winged angles of the style. Although there are 20 similar filaments, not more than 10 (sometimes only 5) bear anthers, the summits of the others being crumpled round the termination of their nerve. We may thus believe in the statement of Palisot on this point, which many botanists have doubted. The disk which supports the floral parts is annular, narrow, prominently erect, crenu- lated on the margin, and placed around the deeply hollow vertex of the ovary, wherein the stigma is ensconced; the latter, peltately supported by a short, broadly 5-angular style, is large, depressed and pentagonal, hollow in the centre, with 5 radiating furrows on the flat summit, each terminating near the angles, in a gaping gland filled with stigmatie tissue. The calyx consists of 5 equal, triangular sepals, which surround the disk, are valvate in sestivation, but afterwards quite rotate; they are thick, flat, with square margins, which end externally in a short mucronate point, on each side of which there is generally a distinct imbedded gland. 1 Seeing them upon the second whorl, he says, ** But just at the bend they often present a slight exerescence or fold, representing perhaps the traces of an anther” (Journ. Linn. Soc. vol. x. p. 495). B2 4 MR. 7. МТЕК ON NAPOLEONA, The inferior ovary is shortly turbinate, varying in the number of its cells. Beauvais figures about 20 radiating cells in his typical species. Jussieu, who reexamined the same plant, does not contradict this, but says he found four ovules in each of its cells. In N. Heudelotii the ovules are superposed in pairs. In М. Whitfieldii Dr. Lindley found the ovary to be 5-celled, with only 2 collateral ovules in each cell. In №. Mann Isaw several ovules in each cell superposed in pairs. One curious point of structure was noticed by Dr. Lindley in the ovary of №. Whitfieldii which had been preserved in spirits: the axis, round which the cells radiate, was hollow, in continuation of the hollow style; and an open foramen was seen leading out of that space into each cell, above the point of suspension of the ovules. He figures this in fig. 5 (Veg. Kingd. p. 728); and he describes, as shown in fig. 6, a depressed dark spot on each side of the ovule, which I have witnessed in another species—a circumstance to which I shall have to refer presently. The fruit is subglobular, about the size of a small apple, deeply umbilicate at the summit, within which cavity are seen the persistent sepals, disk, style and stigma; it is indehiscent, with а coriaceous pericarp, not always very thick, and varying in the number of its cells. The seeds are solitary, or 2 superposed in each cell, sometimes enveloped in a thin pulp, sometimes without pulp; they are obconically globose, and subangular by com- pression, or they are oblong, reniform, laterally compressed, exhibiting generally, in the deep ventral sinus, a broad oblong cicatrix, like that in the seeds of many Sapotacea, showing where they are agglutinated to the inner angle of the cell and to the sides of the dissepiments—an important feature not hitherto observed; the testa is thin and membranaceous, generally evanescent. Тһе embryo consists of 2 equal fleshy plano- convex cotyledons, with a short radicle, which is hidden between them at the lateral sinus, where also a large centrifugal plumule lies concealed. The several species described by authors are reduced to two by Messrs. Bentham and Hooker (Gen. Pl. i. р. 724), but they are amalgamated into one only by Prof. Lawson (Oliv. Afr. Fl. p. 439). It appears to me, on the other hand, from the characters given by different authorities, and from others indicated by the specimens in our herbaria, that several species may easily be recognized by the following data. Тһе typical species comes from the kingdom of Benin, at Waree, on the river Escardos, where, in alluvial ground, it forms а small tree, 7 or 8 feet high. Тһе species found by Heudelot in the high land of Foota Jalhoo, 1000 feet above the level of the sea, and 500 miles in the interior of the Senegal coast, is a tree with a straight trunk 25 to 33 feet high. Most of the other species grow near the sea-coast. Vogel’s species has been found at the two extremities of the Gold Coast, at Cape Palmas, by Vogel, and at Lagos by Baxter. Whitfield’s specimens were found by him at Sierra Leone and the river Nuñez, where it forms a shrub with the appearance of a Camellia, 5 to 7 feet in height. Mann’s plant from Fernando Ро is a tree 20 to 25 feet high. Mann’s specimens from Old Calabar are very different, and form a smaller tree, only 10 to 15 feet high. Welwitsch states that his plants from the interior of Angola were obtained from a tree with a simple straight trunk, 12 feet high. The leaves in Beauvais's plant are oblong, constricted at the summit into an obtusely OMPHALOCARPUM, AND ASTERANTHOS. 5 narrow acumen, at the bottom of which, on each side, is а broad tooth; they are acute and biglandular at the base, upon a very short petiole. Іп Heudelot’s specimens they are narrower, subacute at the base, terminate at the summit in a simple narrow obtuse point, and they have a slender petiole. In Vogel’s specimens they are more elliptie, with a narrow mucronate acumen, are more coriaceous, with undulating cartila- ginous margins, where several dark spots are seen, indicating as many abortive teeth ; they are pale above, yellowish beneath, biglandular near the short fuscous petiole, and have about 9 pairs of ascending nerves. Іп Whitfield's numerous specimens the leaves are longer, with a long, narrow acumen, are of a darkish green colour above, and of a peculiar lurid brown beneath, are subcoriaceous, with about 8 pairs of more patent nerves, rarely with basal glands, and have longer petioles. In Mann's plant, from Old Calabar, the leaves are nearly double the size of any of the preceding, are more oblong, suddenly contracted at the apex into a very long narrow acumen, below which is seen an obtuse tooth on each side, as in Beauvais's plant, but with the addition within it of a conspicuous gland; they are acute at base, with 2 glands near a very short dark-red thicker petiole; the margins are undulated and cartilaginous ; and they have ten pairs of straight ascending nerves. Іп Mann's plant from Fernando Po the leaves are still longer (10 inches), broader at the base, terminating abruptly at the summit in a short simple obtuse acumen; they are more coriaceous, very pale, with an almost encaustic surface above, pale yellow, and subnitid beneath, rarely with basal glands, with 12 pairs of prominent diverging nerves, and short stiff fuscous petioles. In Welwitsch's plant from Angola the leaves are of median size, much more membranaceous, more elliptic, of a more opaque green on both sides, obtusely and distinetly toothed on their thin margins, without basal glands, with about 7 pairs of subascending nerves, have extremely short petioles, with their margins subdecurrent with the angles of the branches. This constant difference in the leaves and habits of the plants is accompanied by equally marked peculiarities in the flowers and fruits, indicating the distinctness of the species. In the flowers the number of sepals is constantly 5 (except where, as rarely happens, they are reduced to four by abortion); the number of marginal lobes and rays of the corolla, the number of segments in the several whorls of the corona, vary in the several species, but they are always multiples of 5. The following table gives in columns the number of parts in the several whorls of the flower, and the relative dimensions of the calyx and corolla in the several species :— Corona. Relative diam. Lobes and of calyx and Species. anat p TR corolla in lines. Corola: | Outer whorl. | Second whorl.| "ner Whorl | д | of filaments. Calyx. | Corolla, М. imperialis ........ 35 35 35 20 10 8 18 Ni Heudeloüi ..........: 40 40 90 10 7 15 RO nM 2.6: 30 75 40 90 5 8 90 N. Whitfeldi ...... 35 60 50 20 5 9 18 N. euspidata ........ 40 7 40 20 10 21 AR МЕНИИ se. 30 40 40 20 11 13 | №. angolensis ........ 30 60 45 20 10 12. | 21 6 MR. J. MIERS ON NAPOLEONA, Those botanists who have advocated the amalgamation of the species, place little reliance on the colour of the flowers, which they say is variable; but Beauvais, Jussieu, Whitfield, and Welwitsch assert that the colours are constant; and this appears con- firmed by the dried specimens. In Napoleona imperialis the corolla is of a beautiful ultramarine blue, the corona of a roseate colour. In NW. Hewdelotii the corolla is of a reddish purple, the corona of a pale rose-colour. In №. Vogelii the corolla is yellow, with a tinge of red about the base, the corona being yellowish white. In N. Whitfieldii the eorolla is of a uniform apricot-colour for more than half its length, then suddenly red towards the base, the corona is yellow, the filaments of a rich crimson colour. In N. Mannii the corolla is only half ihe usual size; this, as well as the corona, is of a pale yellow, the filaments quite white. In М. cuspidata the corolla in the dried state appears of the same colours as in №. Whitfieldii, but the corona is of a deeper roseate hue. In М. angolensis the corolla is white, the corona is white, but the campanular base of the 2nd whorl is of a violet apricot-colour, and the staminal whorl of the same hue. The fruit varies in size in the several species. In М. imperialis it is said to be spherical, with a softish pericarp about the size of a small apple, with many seeds imbedded in fleshy pulp. In №. Heudeloti it is figured as spherical, has а soft coriaceous pericarp 1 or 2 lines in thickness, with 5 thin persistent dissepiments, each cell containing 4 seeds in superposed pairs, of a shape very different from all others. In N. Vogelii the fruit is globular, somewhat depressed, obsoletely 5-lobed in the periphery; the pericarp is 24 inches in diameter, and 12 line in thickness, externally of a dark maroon colour, with pale spots, is about 10-celled, contains about 12 seeds, some of them slant- ingly superposed; and there is no appearance of pulp in the specimen preserved in spirits. In N. Whitfieldii the fruit is 24 in. in diameter, of a more depressed globular form, with a pericarp not thicker than a common card, of a pale rose-colour, with white spots, is very brittle when dry: it has apparently been 5-celled ; but no remains of disse- piments were visible in the specimen in the British Museum. Whitfield says that the seeds are enveloped in an edible pulp. In М. angolensis the fruit much resembles that of the preceding in size and colour, but the pericarp is somewhat stouter, more firm, obsoletely 5-lobed on the periphery; in the dry state it seems to have been 5-celled, and the seeds have probably been enveloped in pulp. These many differential characters, though long disregarded, are positive, and fully justify the maintenance of several good species. Up to 1845, indeed, three of them were acknowledged, when in that year Dr. Lindley proposed a fourth, upon Whitfield’s specimens. In 1848, however, Sir Wm. Hooker, in figuring and describing the latter from a living plant, pronounced decidedly that it did not differ specifically from the original species, .N. imperialis, or from Heudelot’s plant; for he could discern no dissimi- larity between the three plants, except that of colour, which was due solely to their fading in drying. This broad assertion so puzzled Van Houtte, that he wrote immediately to Prof. Decaisne, asking him to examine the typical specimens of Beauvais and Heudelot’s plants in the Paris collections, in order to determine this point. The learned Professor ' Flore des Serres, vol. iv., sub tab. 386, 387 (1848). OMPHALOCARPUM, AND ASTERANTHOS, 7 replied, that he had done so, and could not understand upon what ground the convic- tion of Sir Wm. Hooker rested, as the latter had evidently forgotten that the plants of the two French botanists were examined by them in their growing state, in localities far apart, the blue colour of the one and the purple of the other being noted by them on the spot (and not in the dried state), specifying the time of year they were thus seen in flower. М. Decaisne added that Sir Wm. had failed to notice several facts, such as the peeuliar glands in the plants, the difference in the size of the fruits, the great dissimi- larity in the seeds, the comparative size of the plants, and other characters; he con- cluded that, until he saw proof to the contrary, he should continue to regard Heudelot's | plant as a species widely different from Whitfield’s. Не could not speak with equal certainty in regard to N. imperialis and N. Heudelotii ; but Jussieu, in comparing these two plants, speaks decisively on this point’, stating that their differences are so great that it is impossible to mistake one for the other. In the following year(1849) Mr. Bentham supported the view of Sir Wm. Hooker, and even went so far as to include N. Vogelii in the same category, thus amalgamating the four species into a single one, N. imperialis? ; but in 1865 he somewhat modified this opinion by admitting the existence of two species *. Prof. Lawson afterwards reduced all the plants of Napoleona to a single species*. Sinee that time little fresh evidence has been offered on the subject, except the valuable information contributed by Dr. Masters in his excellent analysis of the living flower of Whitfield's plant?, when he compared this evidence with the published details of other botanists. He pointed out many of the notable differences I have enumerated in the pre- ceding pages, not only in regard to the flowers, but dissimilitude in the habit, in the leaves, fruit, and seeds; still he drew no inferences from them as applicable to the determination of the species; on the contrary, he says, “ si all these plants belong to one variable species, or whether there are four, ог. five distinct forms, is a question no one can answer at present " 5. Notwithstanding this, as the differences are so many, so well marked, and so constant, I have ventured: to employ them in the main- tenance of the following species :~- 1. NAPOLEONA IMPERIALIS, Pal. Beauv. (non Hook.) Fl. Owar. ii. p. 32, tab. 78; DC. Prodr. vii. p.530; Lam. Encl. Suppl. iv. p. 57; Dict. Se. Nat. tab. 66; Reich. Exot. Bot. tab. 187; Sertum Bot. (Brux. 1828), tab. 4.; Van Houtte, Flor. des Serres, i. p. 1, tab. 1; Walp. Bot. Reg. vi. p. 408: arbuscula, ramulis tenuiter teretibus, glabris: foliis oblongo-ovatis, apice repente constrictis, et in acumen lineare ob- tusulum productis atque hie utroque latere dente obtuso munitis, cæteroquin integris, marginibus саг асіпеів, supra viridibus, costa plana, nervis divergentibus utrinque 7, subtus fuscioribus, costa nervisque flavidis prominentibus, imo 2-glan- dulosis, petiolo limbo 30—-35plo breviore : floribus axillaribus, 2—3, fasciculatis, sessi- libus, quorum unico simul expanso cum 2 alabastris imo pluribraeteolatis, bracteolis rotundatis; sepalis 5, acutis, apice extus 2-glandulosis ; corolla orbiculata, rotata, ! Ann. Sc. Nat. 2"* sér, п. 226. * Niger Flora, p. 361. * Benth. & Hook. Gen. Pl. i. p. 724. * Oliver's Afr. Flor. p. 489. * Proc. Linn. Soc. x. p. 492. * Loc. cit. p. 494. 8 MR. J. MIERS ON NAPOLEONA, pulchre azurea, rotata, margine crenulato, subtus nervis crassis subulatis radiantibus 35 donata; corona rosea, paulo minore, verticillo primo e segmentis 35 linearibus rotatis liberis, verticillo secundo e segmentis 35 linearibus imo in cupulam con- matis, superne liberis, et centrum versus inflexis, verticillo interiore e filamentis 20 liberis, plano-linearibus, crebre uniseriatis, inflexis, quorum 5 antheriferis; pistillo generis: drupa globosa, apice umbilicata, sepalis stigmateque coronata, pluriloculari, plurisperma, seminibus in pulpa nidulantibus. In regione Benin, ad Waree: v. в. in herb. Hook. (Beauvais). The above details are derived from the description of Beauvais, and from the incom- plete specimen at Kew sent from the Paris Museum. The species is abundant in woods behind the royal residence on the river Escardos, growing to the height of 7 or 8 feet. The leaves are 43-6 in. long, 2-22 in. broad, on a petiole 2 lines long: the calyx ex- panded is 8 lines in diam. ; the corolla is 18 lines in diameter. 2. NAPOLEONA HEUDELOTII, Juss. Ann. Өс. Nat. 2™° sér. ii. 227, tab. 4; Fl. des Serr. i. р. 2, tab. 2; Ann. Hort. (1822) tab. 10; Walp. Rep. vi. 408 : Napoleona imperialis, var. purpurea, DC. Prodr. vii. 750: arbor ramis verticillatim horizontalibus, ramulis subtenuiter teretibus, subangulatis, brunneis, striolatis: foliis oblongo-ellipticis, imo subacutis, apice in acumen breve obtusulum sensim aut repente constrictis, margine integris, tenuiter chartaceis, utrinque pallide viridibus, opacis, et minute granulatis, subtus vix obscurioribus, nervis tenuibus supra immersis, subtus nervis pallidis venisque prominentibus, imo 2-glandulosis, petiolo fusco limbo 20plo breviore : inflorescentia przecedentis ; corolla ejusdem formee sed purpurea, margine crenata, nervis subulatis 40 radiantibus; corona rosea, verticillo exteriore e seg- mentis linearibus liberis 40 nervis corolle alternis, verticillo 2% imo poculiformi, margine in lacinias 40 incurvas fisso ; filamentis 20, antheris 10; stylo et stigmate generis; ovario infero, 5-loculari, ovulis in quoque loculo 4, collateraliter super- positis: drupa globosa, ambitu obsolete 5-loba, 5-loculari; seminibus in quoque loculo 4, superpositis et collateralibus, pulpa involutis, angulato-globosis, lateribus compressis, apice subcuneatis, emarginatis, et hine ad axem affixis, testa tenui; coty- ledonibus plano-convexiusculis, crassis, carnosis ; radicula in sinu immersa, ad axem spectante. Іп Senegambia interiore: v. s. in hb. Mus. Brit. et Hook. Foota Jalhoo (-Heudelot). The Museum specimen was sent by Jussieu to Mr. Robert Brown, the other one to Kew, from the Paris collection, both collected by Heudelot under the circumstances mentioned in a preceding page. It forms a tree, which is rare, with a straight trunk about 30 feet high, growing in freshwater places on the elevated tableland of Foota Jalhoo, in ferruginous rocks. The axils of its slender branchlets are 3-1 in. apart; the leaves are 33—55 in. long, 14-2 in. broad, on petioles 2 lines long: of the three axillary buds, only a single flower is developed at the same time. Тһе corolla, of a reddish purple colour, is 14 in. in diameter; the corona, of a rose-colour, is half that diameter. Тһе drupe, marked with white spots, is 12 in. in diameter, 14 in. high, the coriaceous pericarp being 1i-2 lines thick; it is 5-celled, containing 20 superposed seeds, which are 6-8 lines in PA] OMPHALOCARPUM, AND ASTERANTHOS. 9 thickness, roundish, somewhat flattened on the sides by pressure, a little narrowed towards the point of their attachment at the axis, near which the radicle is embedded in the very fleshy cotyledons. ` З. NAPOLEONA Үовкілі, Hook. et Planch. in Hook. Icon. tab. 799; Benth. in Niger Flora, p. 360, tab. 49; Walp. Ann. i. p. 475: ramulis tenuibus, pallide brunneis, sulcatis; foliis ellipticis, imo acutis, apice in acumen longum angustum ѕере callo- sum repente constrictis, subundulatis, ad marginem cartilagineum punctis fuscis obsolete serratis, chartaceis, supra pallide viridibus, opacis, nervis tenuibus semi- immersis, subtus fulvescentibus, opacis, nervis venisque prominentibus ; sub lente crebre impresso-punctulatis, imo petiolum versus biglandulosis ; petiolo tenui, fusco, subtus corruguloso, limbo 18plo breviore : inflorescentia consueta ; sepalis 5, rarius 4, apice biglandulosis; corolla orbiculata, depresse concava, pro dimidia parte externa flava, interne sensim rubra, nervis 30 acute lanceolatis radiantibus extus granulato- pruinosis munita; coronæ flavidze verticillo 1° e segmentis 75, anguste linearibus, liberis, radiantibus, vertieillo 2° e segmentis 40 linearibus, pro dimidia parte in cupulam coalitis, marginem versus liberis et incurvatis; filamentis 20, consuetudine liberis et incurvis, antheris fuscis 10; ovario pluriloculari, ovulis perplurimis super- positis : fructu depresse subgloboso, ambitu obsolete 5-lobo, calyce coronato, fusco- purpureo, flavide maculato, 7-11-loculari; seminibus 12 vel pluribus in loculis solitariis, vel oblique superpositis, axi affixis, oblongis, compressis, ventre reniformi- sinuatis. Іп Africa occidentali: v. pl. s. in hb. Hook. Cape Palmas (Vogel), Lagos (Baxter, 20144) ; v. fr. in Mus. Kew. in spiritu (Vogel), siccum Sierra Leone (Lee). Baxter's plant may prove to be distinct; but as it greatly resembles Vogel's in its cha- racters, I have considered it identical. The acumen in the leaves of the former is longer and narrower, the leaves are longer and more acute at the base; and the two come from localities widely apart. Тһе leaves are 31—4 in. (іп Baxter’s 5 in.) long, 13-2 in. broad, on petioles 11-2 lines long. Тһе corolla is 20 lin. in diam.; when not expanded, and only half that size, it is of a pale yellow colour ; when full-grown it is a bright red colour towards the base. The fruit is 2} in. in diam., 2 in. high, and is well represented in Hook. Теоп., though somewhat larger than the specimens; there is no appearance of pulp, either when preserved in spirits or when dry ; the dissepiments are more persistent than in any other species, where they are probably membranaceous and disappear with the pulp. 4. NAPOLEONA WHITFIELDII, Lindl. Gard. Chron. 1844, p. 780; Bot. Reg. vol. xxx. Misc. p. 77; Veg. Kingd. p. 728 cum icone; Van Houtte, Fl. des Serres, tab. 386, 387; Decaisne, Rev. Hort. (1853), p. 301, tab. 16; Schnizl. Iconog. ii. tab. 159. fig. 1 ad 8: Napoleona imperialis, Hook. (non Palis.), Bot. Mag. tab. 4387: ramulis verti- cillatis, patentissimis, pallide brunneis, rugulosis: foliis oblongo-ellipticis, imo sub- acutis, apice in acumen lineare obtusulum constrictis, integris, marginibus sub- cartilagineis, rigidule chartaceis, supra pallidissimis, ssepe encausto-canescentibus, valde opacis, nervis paullo prominulis, subtus obseurioribus, brunnescentibus, opacis, SECOND SERIES.—BOTANY, VOL. I. б 10 | MR. J. MIERS ON NAPOLEONA, sub lente scabridule lepidotis et punctulatis, costa pallida nervisque prominulis, ad basin rarius 2-glandulosis ; petiolo fusco, canaliculato, limbo 16plo breviore : floribus axillaribus generis ; corolla rotata, subconcava, peripherio armeniaca, centro coccinea, subtus costis planatis 35 viridescentibus et scabridulis sensim acutis, margine in lobulos totidem crebre plicatulos incisa ; coronze соссіпеге series externa e segmentis 60 liberis anguste linearibus constat, series intermedia e segmentis 50 linearibus ad medium campanulatim coalitis apice liberis et incurvatis, series interna e filamentis 20 linearibus introflexis, imo in annulum brevissimum connatis, et hie ad discum elevatum crenulatum affixis; antheris 10; ovario 5-loculari, ovulis 4—6 in quoque loculo collateraliter axi superfixis: drupa depresse globosa, pallide rosea, aibide maculata, 5-loculari, pericarpio tenuissimo fragili; seminibus paucis, oblongis, com- pressis, ventre reniformibus, in pulpa mucilaginosa eduli immersis. In Africa tropica : v. в. pl. in herb. Mus. Brit. Sierra Leone (Whitfield), Sierra Leone (LAfzelius) ; in hb. Hook. Sierra Leone et Rio Nuñez (Whitfield): v. fr. s. in Mus. Brit. (Whitfield). Whitfield described this plant as like a camellia-bush ; and Lindley says it has a softish wood remarkable for its large medullary rays, with abundance of spotted vessels inter- mingled with acicular tubes of woody tissue, like those found in the germinating radicle of Rhizophora. In Whitfield’s specimen from Rio Nufiez, the branchlets spread at right angles; and Sir Wm. Hooker figures them as verticillated and horizontal. Тһе axils generally are 2-15 in. apart; the leaves are 42-62 in. long, 13—21 in. broad, on petioles 3—4 lines long, near which below a gland is seen on each side; they are furnished with about 8 pairs of divergent nerves. The inflorescence, as in most of the other species, consists of 3 fasciculated flowers, almost sessile, on extremely short pedicels, which bear 4 small roundish decussately imbricated bracts, each 2-glandular. Of these 3 flowers only one is at first developed, the other two in the meanwhile remaining in the bud- state: the sepals are square on the edges, gradually narrowing into a wedge-shaped point, below which externally are 2 imbedded glands; they are 3 lines long, 3 lines broad at the base, and are marked outside with many yellowish prominent granules ; they spread rotately to a diameter of 9 lines on the summit of the inferior turbinate ovary. The corolla is 11 in. in diameter, The fruit is 21 in. broad, 14 in. high, crowned by the per- sistent calyx and stigma, and there umbilicated; it differs from all other species in its pericarp not being thicker than a thin card, in being very brittle; it shows internally the slight vestiges of 5 dissepiments, which would seem to have been membranaceous ; for no traces of them are seen in the dried fruit, nor of the edible mucilaginous pulp in which the seeds are said to be enveloped. There are about 12 seeds, laterally compressed, 12 lines long, 7 lines broad, and 43 lines thick: the thin membranaceous testa decays in drying, leaving the embryo bare, except on the cicatrix on the ventral sinus, which is seen often on one side only, where the seeds have been collateral in each cell, and where they have been there attached on one side of the dissepiment. 5. NAPOLEONA CUSPIDATA, nob.: ramulis subvalidis, pallidis, costato-striatis: foliis ma- joribus, oblongis, imo obtusis aut subacutis, apice in acumen lineare longum glan- duloso-mucronatum subito constrictis, marginibus undulato-erispatis, integris, sub «diac PARA iR OMPHALOCARPUM, AND ASTERANTHOS. 11 acumen dente magno obtuso medio subtus glandulifero utrinque munitis, sub- coriaceis, pallidissimis, opacis, cum nervis tenuibus rectiuscule adscendentibus utrin- que circa 10, venis transversis reticulatis, subtus fulvescentibus, subnitentibus, costa nervis venisque prominentibus, ad basin breviter revolutis et hie glandula parva notatis; petiolo semitereti, substriato, fusce rubido, limbo 60plo breviore : floribus 3, sessilibus, quorum 2 alabastris, et unico expanso, e gemma axillari imo bracteis 4 parvis rotundis decussatis biglandulosis munita ortis; sepalis 5, acutis, apice glan- dulis 2 signatis; corolla orbiculata, magna, extus nervis 40 lineari-lanceolatis, mar- gine in lobos totidem breves crispatos secta, imo purpurea, marginem versus dilute flavida; coron:e verticillo exteriore in lacinias 70 radiantes diviso, his anguste line- aribus, flavidis, e basi omnino liberis, quam corolla subbrevioribus, singulis vesicula parva effoeta paullo supra basin extus munitis, verticillo intermedio depresse globoso, 40-nervio, nervis paullo supra basin vesieulis totidem parvis extus donatis, margine in lacinias 40 triplo breviores lineares obtusas inflexim conniventes scisso, verti- cillo interno e laciniis 20, a basi liberis, late linearibus, corollæ fere sequilongis, ‘primum paullo erectis, versus medium inflexim descendentibus, apicibus obtusis sub stigma conniventibus, et hie anthera fertili oblonga fusca singulatim munitis. In Africa tropica: v. s. in herb. Hook, Old Calabar (Mann, 2272). This plant, together with N. Manil has been referred to №. Vogelii, but differs from it in many salient characters. The axils of its thickish branchlets are 1 in. apart; the leaves are 6-81 in. long (including the acumen 1 in. long), 2-22 in. broad, on а petiole 1-9 lines long. The calyx expanded is 10 lines across, its thick sepals being 3 lines long, 21 lines broad at their base; the corolla is 14 in. in diameter; its nerves, not very pro- minent below, touch each other at their base, terminating acutely in the middle of the peripheral lobes. The 70 segments of the external whorl of the corona are 6 lines long, lline broad, their prominent effete vesicles forming below a conspicuous moniliform ring 1 line beyond their base: the 40 segments of the second whorl, if free, would be as long as those of the external whorl; but they are confluent for two thirds of their length into a depressed globe, their free portions connivent at the centre, and the underpart of this globe not far from its base shows a prominent moniliform ring of 40 vesicles, similar to those of the outer whorl: the 20 free segments of the inner whorl are 7 lines long, 1 line broad, incurved as above mentioned, each furnished extrorsely near its apex with an adnate oblong fertile anther, these all crowding round the stigma. The style is ex- tremely short and pentagonoid; the peltate pentagonal stigma, 5 lines broad, has 5 glan- dular spots near its angles, and scarcely rises above the level of the disk; the inferior ovary is compressedly turbinate, 5-celled, with several ovules in each cell, radiately attached to the projecting placentze оп the axis. Тһе fruit is unknown. 6. NAPOLEONA MANNII, nob.: ramulis subvalidis, pallidis, suleato-striatis, epidermide resiliente: foliis majoribus, oblongis, imo obtusis vel rotundatis, apice in acumen subbreve obtusulum repente constrietis, marginibus undique integerrimis, sub- undulatis, rigide chartaceis, supra pallidis, valde opacis, seepe cano-encaustis, nervis longe adscendentibus, subimmersis, subtus pallide fulvis, 2. minute C 12 МВ. J. MIERS ОМ NAPOLEONA, impresso-punctulatis, costa nervisque prominentibus, venis transversis et prominulis, imo seepius eglandulosis ; petiolo fusco, canaliculato, ruguloso, limbo 48plo breviore : flore solitario, breviter pedicellato, e gemma bracteolata axillari orto, pro genere parvo; sepalis acutis, sub apicem biglandulosis, extus pallide viridibus, pruinosis, intus flavidis; corolla undique flava, extus pruinosa, cum nervis obsolete costæ- formibus circa 30, in lobos totidem marginales rotundatos desinentibus, ambitu sepalos vix excedente; coron: alb: verticillo exteriore e laciniis 40 anguste line- aribus, e basi liberis, corollae eequilongis, imo annulo moniliformi vesicularum signato; verticillo intermedio imo cupulari et bic annulo monilformi munito, pro tertia parte in lacinias 40 angustas incurvas centro conniventes diviso; verti- сШо interiore in lacinias 20 lineares complete introflexas secto, apicibus anthera oblonga fertili munitis, cirea stylum conniventibus ; stylo et stigmate ut in charac- tere generis. Іп insula Fernando Ро: v. pl. sic. in hb. Hook. loc. cit. (Mann, 590). А species differing from all others in its much larger and more rigid leaves with a very pale encaustie surface, in its solitary flowers, only half the size of any of the others, of a uniformly pale colour, furnished with a white corona peculiar in the number of its parts. It forms a tree 20-25 feet high, with branchlets 14 line thick, with axils 1 in. apart; its leaves are 10 in. long, 4 in. broad, on a petiole 22 lines long. Тһе pedicel of its single flower is $ line long; the acute sepals are 4 lines long, 3 lines broad at their base; the corolla is 13 lines in diameter. The segments of the first whorl of the corona spread to a diameter of 13 lines; the diameter of the globular contour of the second whorl Slines. The fruit is unknown. 7. NAPOLEONA ANGOLENSIS, Welw. MSS. : ramulis teneris, pallidis, 4-angulatis, rugulosis : foliis elliptico-oblongis, imo subobtusis, apice in acumen subbreve vel longius obtusulum subito constrictis, marginibus distincte et undique crenato-serratis, mem- branaceo-chartaceis, supra viridibus, subopacis, nervis divergentibus semiimmersis, subtus pallidioribus, costa nervis venisque reticulatis prominulis; petiolo supra eanalieulato, subtus valde corrugato, limbo 30plo breviore: floribus in axillis 3, e gemma ortis, quorum 2 alabastris, 1 expanso; pedicello vix ullo, imo decussatim bracteolato; sepalis 4 vel 5, acutis, apice 2-glandulosis, extus viridibus, intus fla- vescentibus; corolla late orbiculari, membranacea, subtus nervis 30 subulatis pro- minulis pallide pruinosis signata, ambitu in lobulos totidem rotundatos lacinulatos incisa, (sec. cl. Welw.) omnino nivea (sicca flavescente) ; coron: verticillo exteriore e laciniis 60 liberis albis, anguste linearibus, acutis, imo tenuioribus, corolla dimidio brevioribus, singulis subtus vesicula minima munitis; verticillo intermedio pro dimidia parte cupuliformi, extus annulo moniliformi vesieularum instructo, margine in lacinias lineares 45 inciso, his introflexis, centro conniventibus; verticillo interiore e laciniis 20 late linearibus, liberis, introflexis, apicibus anthera fertili donatis, stylum circumstantibus ; drupa sicca, depresse globosa, sinuato- 5-sulcata, pericarpio tenui, coriaceo, fragili, pallide rubidulo, punctis parvis flavidis immersis maculato, sicco uniloculari, intus septorum 5 vestigiis signato ; seminibus reniformi-oblongis, subcom- _ pressis. In Angola: v. pl. s. in hb. Welwitsch; v. fr. s. т Mus. Brit. (Welwitsch). OMPHALOCARPUM, AND ASTERANTHOS. | 13 А very distinct species, marked by peculiar characters. It forms a slender tree, 12 feet high, with a trunk 1-21 in. thick, and with slender branchlets; its leaves, remarkable for their serrated margins, are generally 5 in. long, 2-23 in. broad, with about 7 pairs of nerves. Flower generally solitary, about the size of that of the typieal species: the calyx expanded is 1 in. across from point to point; the sepals 4 lines long, 3 lines broad at their insertion round the disk, which is 3 lines in diameter; the corolla is 12 in. in width, or 9 lines long from the line of its attachment to the circumference : the segments of the outer whorl of the corona are 43 lines long, $ line broad; the total length of the half- confluent segments of the intermediate whorl is 8 lines, the free portion being one half of this, or 4 lines; the segments of the inner whorl are 6 lines long, 1 line broad: the fertile anthers 2 lines long. Тһе fruit is 21 in. in diameter, 13 in. high: the seeds are 1 т. long, 7 lines broad, 4j lines thick; they are of а pale red colour, with close yellow immersed glands; the transverse short radicle is completely embedded within the sub- stance of the cotyledons, close to the ventral sinus, which have no chink there, as in N. Whitfieldii, The tree was found in dense woods at Quitapo (Omnengue), between Ndelle and Rio Zuinha, in flower in May 1856, and in fruit at the end of June of the same year. After this critical examination of the several species, we can perceive nothing, either in the floral or carpological structure of Napoleona, that bears the slightest analogy to the Myrtacee ; it is equally irreconcilable with the Barringtoniee and the Lecythidacec. In order to trace its real affinity, we will compare with it Omphalocarpum, a genus derived from the same vicinity. 9. OMPHALOCARPUM. This genus was first discovered by Palisot de Beauvais in 1787, in the same province "where he found Napoleona; but it was not described and figured till 1804', when he pointed out its affinity with Sapotacee. АП botanists since that time have accorded in this view °; but in 1862 Messrs. Bentham and Hooker, without adducing reasons, placed the genus in Terastremiacec?, and subsequently Prof. Oliver followed their example‘; but notwithstanding such high authorities, I venture to think there is little in its strueture conformable to that family, and an overwhelming amount of evidence to show that it is truly Sapotaceous. It forms lofty trees, with erect trunks; the leaves, not large, are approximated at the ends of the branches; the inflorescence generally issues out of the main trunk, or appears on the bare branches, rarely in the axils; the flowers, 3 to 6, usually spring from bracteolated tufts, and are mostly pedicellated, but in one species they are sessile; they are not very large, and are perhaps hermaphrodite in the typical species, but in the others they are more or less polygamous, as in other Sapotacee?. Of this we have parallel examples in that family where there is an г FL Owar. i. р. 6, tab. 5 et 6. * Lam. Dict. Suppl. iv. 140, tab, 966 ; Brown, Prodr. p. 529; De. Prodr. viii. 207; Endl. Gen. p. 741; Lindl. Veg. Kingd. р. 591; Hook, Niger Flora, р. 441. 3 Gen. Plant. i. 185. * Flor. Afr. 171. * Miq. Monogr, Sapot. in Mart. Fl, Bras. fasc. xxxii. p. 38, where he says “ flores fere semper hermaphroditi." - 14 MR. J. MIERS ОМ NAPOLEONA, exuberant production of flowers, as in Mimusops. The sepals are 5, roundish, subcori- aceous, very imbricated, the 2 exterior nearly concealing the others in sestivation, as in Зароасее. The corolla is described by DeCandolle’ as gamopetalous and infundibuli- form, deeply cleft into 6 or 7 distinct segments : this, however, is not a correct definition ; for in reality there are 5 petals (in one species more), all furnished with cuneiform claws, which are firmly agglutinated upon the outer surface of a funnel-shaped thin disk, leaving the upper and broader portions apparently as so many free segments, which are erect, oblong, and somewhat imbricated on the margin: the proof of this is clearly shown on examining the outside of the tube, where we see the margins of the claws free, like so many very narrow wings, extending to the base, leaving narrow portions of the bare disk visible between them. The disk, as above shown, is tubular, as long as the claws of the petals, and agglutinated to them, just as occurs in Mimusops and Lucuma; and in like manner it supports upon its margin 5 sets of stamens, nearly the length of the free portion of the petals, and placed opposite to them in a single series; their filaments are straight and slender, very pointed and recurved at the summit, where they support each a versatile anther, composed of 2 long collateral cells, divergent at the base, and fixed extrorsely upon a fleshy connective, which is uncinately excurrent at the apex; if we cut this anther in its young state transversely, it appears 4-celled ; but аз the dehiscence takes place in the deep furrows, it is seen to be only 2-celled, each cell containing 2 parallel masses of cohering pollen grains; after bursting they become separated, then appearing oval and reticulated: the same number of filaments, similar in every respect, are seen in the 4 and % polygamous flowers, but they are antheriferous in the former and bare in the latter. Between these sets of fertile stamens, and alternate with the petals, are seen as many processes, strap-shaped below, widening upwards, where they are flabellately laciniated into about 7 setaceous segments, the middle one longest, attaining the length of the fertile stamens; these exactly resemble the so-called staminodes found in Mimusops, Lucuma, and Dipholis. The ovary is superior; in the с flower it is rather small, orbicular, fixed on a short stipitate support, depressed above, where it is marked by many radiating grooves, bearing in the centre an erect slender style, reaching the anthers, and terminated by a small many-lobed hollow stigma; it contains an irregular number of cells, each with а small ovule, probably all abortive; in the $ flower the ovary is larger, and fixed on a stipitate support, is many-grooved, and expands above gradually into a stout conical coriaceous style (as long as that of the g flower), obtuse at the summit, where it is hollow and many-lobed: when this style is cut transversely, we see there as many distinct channels as there are cells in the ovary, and descending into them; there are about 24 cells, radiating from a hollow central axis, with a single ovule in each cell, horizontally attached : all this agrees with the structure of Mimusops. The fruit is large (44—82 in. in diam., sometimes much larger), of a very depressed orbicular form, radiately suleated, umbilicated in the centre by a large hollow space, in the bottom of which the persistent style is seen : the indehiscent pericarp is 1-1 inch in thickness; the epicarp is hard and subligneous: the mesocarp, composed of many distinct, ovoid, subcompressed, grumous concretions (as in Labatia), like the nodules of a mineral con- * Prodr. viii. 287. OMPHALOCARPUM, AND ASTERANTHOS. 15 glomerate, irregularly interwoven in two series; these nodules are easily separable in the fresh state, but afterwards become solidified and compact: the central space is divided into about 20 narrow cells by dissepiments, which radiate around a large hollow orbiform ligneous axis, which is internally smooth and polished : the placental angle of each cell thus inclines upwards and outwards, and bears a single seed attached to it by a long linear caruncle or scar. This seed is oblong, very compressed, much thinner towards the margins, rounded on three sides, but straight along the whole ventral edge, and there marked by a pale opaque soft caruncula, as in Sapota, where it is called by DeCandolle the umbilieus, but by Gaertner the umbilical area, to distinguish it from the true umbilieus always seen on one of its extremities, as figured in his plate 203, and so described in his text under Lucuma, Sapota, Chrysophyllum, Sideroxylon, Mimusops, and Imbricaria; in the present instance the hilum is seen at its lower extremity, like an oval hole divided by a septum, the remaining part being filled with the ascending raphe, terminating at the opposite extremity in a small foramen, which it perforates to terminate in the apical chalaza. The testa is polished externally, is rather thick, very hard, osseous, dark brown, expanded at each ventral angle into a short hooked rostrum, the lower one marked by a transverse chink (the micropyle); the inner integument is brown, chartaceous, and adheres firmly to the inside of the testa. Іп the only seed I was able to examine, the thinnish albumen alone was present, the enclosed embryo being abortive or decayed; but this is well figured by Palisot, who figures it as con- sisting of 2 large thin foliaceous cotyledons, united at their base by a small inferior radicle. The genus is thus seen to agree remarkably with Sapota in the structure of its flower, of its fruit, and especially in that of its seeds; indeed it differs from the latter only in its large umbilicated fruit with а hard ligneous pericarp, having a large hollow axis. I will here record the two following species. Я 1. OMPHALOCARPUM PROCERUM, Pal. Beauv. Fl. Owar. i. p. 6, tab. 5 et 6; R. Brown, Prodr. p.529; Poir. Dict. Suppl. iv. p. 140, tab. 966; DeCand. Prodr. viii. 208: arbor comosa, trunco altissimo, ramis patentibus, ramulis diffusis: foliis alternis, sessili- bus, lanceolatis, integris, glabris, supra nitidis: floribus solitariis vel pluribus fasciculatis, sessilibus, e trunco nascentibus, hermaphroditis?; sepalis 10 (quorum 4-5 forsan bracteis?), valde imbricatis, ovato-rotundis, concavis, extus sericeo- villosis; petalis 6-7, oblongis, margine undulatis, glabris, imo longe unguiculatis, unguibus ad discum tubulosum perigynum petalis triplo breviorem agglutinatis, eorolla hine pseudo-monopetala ; staminibus in phalangibus 6-7 petalis oppositis margini disci insertis, filamentis tenuibus, antheris oblongis, cum connectivo subu- lato exserto ; staminodiis 6-7, singulatim inter quidque petalum disco insertis, petalis 3plo brevioribus, oblongis, membranaceis, margine fimbriatis ; ovario supero, depresse globoso, multiloculari; stylo subulato-tereti, petalis breviore, persistente; stig- mate capitato, aspero: fructu permagno, indehiscente, orbiculari, valde depresso, plurisuleato, apice profunde umbilicato, extus turgide tuberculato; регісатріо duro, crasso, e concretionibus majusculis confecto, radiatim pluriloculari; semini- 16 MR. 7. MIERS ON NAPOLEONA, bus in loeulis solitariis (pulpa involutis?), oblongis compressis, margine ventrali caruncula lineari signatis, angulis ventralibus singulis rostratis ; embryone oblongo, in albumen tenue subzequilongum immerso, cotyledonibus 2 magnis foliaceis ellip- ticis, radicula parva infera. Іп Africa occidentali (Upper Guinea) ad Buonopozo, regione Galbar ultra regnum Owariense, a cl. auct. lectum (non vidi). According to Palisot, this is а tree with a lofty straight trunk, 60-80 feet high, of which he gives a diminished drawing: the size and exact contour of the leaves are not given, only that they are lanceolate, and therefore different from the specimens in the Kew herbarium from the Camaroons and Old Calabar. Тһе flowers also differ from those of the latter in being hermaphrodite, sessile, and pentamerous; the sepals seri- ceously villous; they are 1 inch long. Тһе fruit is large, often 12 inches in diam.; the seeds are 1 in. long, $ in. broad. From Palisot’s ample generic character, there is nothing that can lead us to suppose that the flowers are otherwise than her- maphrodite. 2. OMPHALOCARPUM ELATUM, nob,: Omphalocarpum procerum, Oliv. (non Palisot), Afr. Flor. i. p. 171: ramulis apice foliiferis, deorsum nudis, axillis approximatis : foliis oblongis, imo longe spathulatis, apice rotundiusculis, vel sensim obtusatis, majusculis, integris, marginibus subrevolutis, rigide chartaceis, glaberrimis, supra rufescenti-viridibus, subpruinoso-opacis, costa planata, griseo-pruinosa, striolata, nervis divergentibus arcuatim nexis, prominulis, subtus ferrugineo-opacis, costa prominente, nervis flexuosis prominulis, venis transversim reticulatis; petiolo brevissimo, late concavo, fusco: floribus 5-6, e nodo bracteolato trunco innascenti- fasciculatis, pedicellatis; sepalis semper 5, ovatis, concavis, pedicello sequilongis, valde imbricatis, quorum 3 exterioribus subglabris, crassioribus, 2 intimis submem- branaceis, cano-tomentellis ; corolla flava structura praecedentis, sed petalis phalangis et staminodiis semper 5: staminibus in quaque phalange uniseriatim 6, petalis oppositis, et iis dimidio brevioribus, erectis, apice subito reflexis; antheris supra medium affixis, extrorsum collateraliter 2-lobis, acute oblongis, conneetivo excur- rente; ovario depresse globoso; stylo subulato; stigmate capitato; in 9 ovario imo breviter stipitato, cylindrico, pluriloculari, stylo cum eo continuo obconice cylin- drieo, stigmate annulari vacuo pluridenticulato dentibus cum styli canalibus ovariique loculis congruentibus: fructu forma precedentis, sed dimidio vel duobus trientibus minore, extus sublzevi (vix tubereulato), 24-loculari; seminibus similibus. In Africa occidentali: v. pl. s. т hb. Hook. d Bagroo river (Mann, 815), Old Calabar (Thomson, 128); 9 Camaroon river (Mann, 719): v. fr. s. in Mus. Кеш. 4 Camaroon river ( Mann). A species very distinct from the preceding; its height is unknown. In Mann's specimen from Bogroo river the leaves are broader, 5-9 in. long, 21-31 in. broad, on - | a petiole 2 lines long; those of Mann's two specimens from the Cameroons are 5-9 in. long, 12-25 in. broad. Thomson’s specimen consists of a single leaf, 93 in. long, 2} in. broad, cuneate at base, where it is } in. broad, without a petiole: it belongs probably to аи ТТЕ а ras cc up ОД, ОТА КГ ГҮ қ meer ? ' iii "m Р И OMPHALOCARPUM, AND ASTERANTHOS. 17 a different species. The flowers are nearly 1 іпеһ long; the sepals are 4 lines long, 3 lines broad; the funnel-shaped disk 4 lines long, the claws of the petals glued to it are of the same length, their free portions above being 7 lines long, 3 lines broad; the filaments in both sexes are slender, erect, 5 lines long; the anthers 2 lines long, acute above, cordate at the base, being fixed behind at the sinus to the recurved apex of the filament, standing extrorsely. The fruit in one case is 4l in., in the other 5} in. in diameter, and 3 in. high, the pericarp being 10 lines in thickness, and composed of con- glomerate nodules, as is well described by Beauvais; the seeds are 13 in. long, 1 in. broad, and 4 lines thick in the middle. It is thus seen that the structure of its flower resembles that of Mimusops, differing only in its isomerous petals, in the greater number of its fertile stamens, and in their greater length; it comes close to Sapota, pu gar in its fruit and seeds; and here it will find its proper place. If we now carry out our proposed comparison between the structures of Napoleona and Omphalocarpwm, we discern many points of singular analogy. They both come from the same region. They have an inflorescence in which the fasciculated flowers grow out of bracteolated nodules, a calyx of 5 sepals, a corolla perfectly monopetalous in one case, pseudo-tubular in the other, both furnished inside with separate phalanges of fertile stamens having extrorse anthers; phalanges of sterile stamens alternately placed in one ease, concentrically disposed in the other ; a plurilocular ovary, with one or two ovules in each cell; an indehiscent plurilocular fruit, orbicular, depressed, and umbilicated in the apex; and what is more remarkable, they both possess the very unusual character of a hollow axis, causing the placentz to become inclined, with a solitary seed in each cell, attached to it by a ventral scar common to all Sapotacee (the * umbilical area” of Gaert- ner), a soft place which is probably only an expansion of the raphe. But, on the other hand, the differences are so great between the two genera as absolutely to prevent the admis- sion of Napoleona into the Sapotacee; it differs in its sepals with valvate sestivation, a completely monopetalous corolla, in having its staminodes arranged in concentric whorls (although Dipholis has phalanges of 3.staminodes arranged concentrically in a double series); it has an epigynous (not a perigynous) disk, an inferior (not a superior) ovary and fruit, exalbuminous seeds (though these occur in Гисита and Labatia). If, then, Napoleona cannot be admitted into the Sapotacee, and there is no other natural order in which it can possibly find a place, it must remain the monotypic representative of a distinct family, the Belvidiacee, as originally maintained by Brown and Lindley, ex- cluding, however, Asteranthos. This family, though now consisting only of a single genus, must be placed in juxtaposition with the Sapotacee ; and it is highly probable that other forms closely allied to it may be found in the forests, so full of rich vegetation, in the yet unexplored regions of tropical Africa. 3. ASTERANTHOS. Although this genus has been associated hitherto with Napoleona, it will be shown that it does not bear the smallest affinity with it. It is entirely of South-American SECOND SERIES.—BOTANY, VOL. I. D 18 MR. J. MIERS ON NAPOLEONA, origin, and, excepting the contour of its corolla, 1 presents quite another aspect, and a floral structure different in every respect. Only one species is recorded. 1. ASTERANTHOS BRASILIENSIS, Desf. Мет. Mus. vi. p. 9, tab. 3; D'C. Prodr. vii. 551. In Brasilia et Venezuela: v. s. in hb. тео et alior. Rio Guainia (Spuce, 3500). The plant above quoted is one of the gems of Dr. Spruce's invaluable collection ; it was found by him on the banks of the river Guainia, the main branch of the Rio Negro, about 60 miles within the confines of Venezuela, and therefore not in Brazilian territory. It is probably a species distinct from Desfontaine's; for the latter has smaller leaves, larger flowers, a broader and deeper calyx, more resembling * the cup of an acorn," has much longer stamens, an ovary said to be inferior, with a 6-rayed vertex, and a stouter style and stigma. Spruce’s plant is apparently a shrub, with upright alternate branches, the axils being 3 in. apart; alternate leaves, oblong, subacute at base, with a long, narrow, obtuse acumen, their undulated margins subrevolute, especially towards the petiole; they are dull above, with nerves wholly immersed, pruinosely opaque and paler beneath, veinless, with a slightly raised midrib ; 23-81 in. long, 1-13 in. broad, on petioles llinelong. Тһе inflorescence consists of a solitary flower in each axil, articulated on a short gemmiform nodule, bearing a minute bract; the pedicel is slender, erect, 9 lines long; the calyx is orbicular, very depressed, 12 line broad, 5 lines in diameter, its con- eave margin cut into about 16 blunt short teeth, each terminating in a curving bristle. The corolla is a circular, depressed, membranaceous cup, plicated in sstivation, as in Cordia; expanded, it is 21 in. in diameter, with an upturned margin, divided into 24 short roundish teeth, sparsely and setosely ciliated on their margins, and they correspond with as many fine nerves radiating from the centre, the intervening spaces being reticu- lated with fine veins; it is of а dull yellow colour when dry; it is attached by its per- forated centre to the margin of a narrow, circular, annular, epigynous disk, 3 lines in diameter. Тһе disk supports very numerous stamens, all quite free from the corolla, in three ог four series of different lengths; the very close slender filaments are straight, the outer series 3 lines, the inner row 2 lines long, each bearing a linear-oblong, 2-celled anther, bursting by lateral sutures, introrsely and dorsally fixed on the curving apex of the filament, a little above the base; the ovary is quite superior, rising within the disk in а сопіса! deeply 8-ribbed vertex, surmounted by a subulate slender style, 3 lines long, terminating in a capitate hollow stigma, with an 8-toothed margin; the ovary has eight cells opposite the ribs, each with а prominent placenta emanating from the axis, and charged with several collateral ovules, arranged in about three superposed series, with the raphes of the anatropous ovules facing each other. Тһе fruit is unknown. This structure is so widely different from that of Napoleona that Asteranthos cannot possibly remain in contiguity with it; the calyx is quite dissimilar in form, and has another xstivation; its flower presents no trace of the corona which forms such a peculiar feature in Napoleona; there is no analogy in the form, structure, or position of . the stamens; the ovary is superior, with a long slender style, and an extremely dissimilar stigma; in fine, there is no single point of resemblance of its parts except in the orbicular shape of its corolla, an adventitious feature of weak value (insufficient to establish any OMPHALOCARPUM, AND ASTERANTHOS. 19 claim to affinity), and even there it is constituted very differently ; for in Asteranthos its texture is that of an ordinary monopetalous corolla, simply marked by several delicate nerves, with intervening reticulated veins, as, for instance, in the campanular corolla of some species of Cordia, which have a very similar zestivation. Searching among the many genera with a monopetalous corolla, we come upon Rhodo- dendron, where we find many unexpected points of resemblance. Asteranthos Ваз pre- cisely the calyx figured by Wight (in his Icones, pl. 1203) of Rhododendron Griffithsi- anum; that species has а broad campanular corolla, with many longitudinal nervures, and is divided on its margin into several teeth; the tube is plicated with a similar æsti- vation; though its stamens are more numerous, they have similar long slender filaments, seated on a thin disk, free from the corolla, and similarly formed anthers; it has a superior conical ovary, of the same shape, marked in like manner by salient ridges, cor- responding with the number of the cells, a similar long slender style, and a remarkably similar stigma. "There would be no difference in the anthers if, in Rhododendron, the apical pores were continued in a slit along the remainder of the sutures, as, indeed, actually occurs in Azalea, where their dehiscence is like that in Asteranthos; other points of analogy are found in the arborescent species of Rhododendron, where the leaves are as frequently alternate as they are opposite, and where the large flowers are, in like manner, seated upon long, slender, solitary pedicels, bracteolated at their base. There is nothing, therefore, to separate this genus from others of the Rhododendree, except a more rotate corolla, with more numerous and shorter lobes. When the fruit is known, other ana- logies may be found to exist; but under our present knowledge of the structure of Aster- anthos, it may safely be regarded as a member of the tribe just mentioned. It may be added that Rhododendron maximum, with a large, almost rotate corolla, is a native of South Carolina, a circumstance which renders the relationship more complete. DESCRIPTION OF THE PLATES. PLATE I. Fig. 1. A leaf of Napoleona imperialis, from a specimen sent by Palisot de Beauvois to Rob. Brown. 2. A leaf of Napoleona Whitfieldii, from Whitfield’s specimen. 3. A fruit of the same species, from Whitfield’s collection. 4, The same in longitudinal section, to show its thin pericarp. 5. A seed from the same, seen on its face. 6. The same, shown on its edge. 7. One of its cotyledons seen on its inside face, showing the radicle and plumule half immersed in it. 8. A leaf of Napoleona angolensis, from Welwitch’s specimens. 9. A fruit of the same species. 10. A longitudinal section of the same, showing its thin pericarp. 11. A seed belonging to the same, shown on its face. 12. The same, seen edgeways. mae 20 Fig. Fig. о C QU к 09 0 ы 5. 15. 16. MR. 7. MIERS ON NAPOLEONA, PLATE II. . А leaf of Napoleona Manni. . A flower of the same, viewed from above. . The same, as seen from beneath. . A leaf of Napoleona cuspidata. . À flower of the same, viewed from above. . The same, as seen from beneath: all nat. size. РтАтЕ ПІ. Section A. Analysis of Napoleona Vogelii. . A flower, seen from above. . The same, viewed sideways. . The corolla detached from its insertion on the disk, and seen from below. . The outer whorl of the corona detached, consisting of 75 free, radiating, narrow segments: ай nat. size. Three of the same segments, shown in different positions: magnified, to show the vesicles or abortive anther-cells, . ‚ The second whorl of the corona, showing 40 segments, confluent on their margin for half their length into a hemispherical cup, while the free portions are inflected and connivent over the stigma. . A side view of the same. | . The same, forcibly expanded, to show the comparative length of the segments. . The inner whorl of the corona, consisting of 10 free stamens, inflected so that the terminal extrorse anthers are brought beneath the stigma. . The same, with their filaments extended to show their relative length: all nat. size. . Three of the stamens, in different positions: magnified. The calyx, seen from below. . 'The same, viewed from above. The stigma: all nat. size. The same: magnified., A longitudinal section of the 5-celled inferior ovary, of the short style and broad stigina, of the abbreviated vertical disk, from which the corolla and the 3 whorls of the corona have been removed, showing also the hollow central axis and 2 erect ovules fixed near the base in each cell : magnified. 17. A diagram of the parts of the flower in longitudinal section: nat. size.. 18. 19. 30. 21. 22. А fruit of the same species, crowned by the calyx, with a quarter of it cut away, showing in longitudinal section the much thicker pericarp, the hollow axis, and 2 seeds in the natural position. А transverse section of the same, showing 9 radiating cells, with a single seed in each cell, without any enveloping pulp, as seen in a specimen preserved in spirits in the Kew Museum. А seed, shown on its face, with a earuncular ventral scar, The same, seen on its edge. The same, on its face, with the thin integument removed. 23. One of the cotyledons, viewed from its inside face, showing the radicle and plumule half-immersed within the sinus, дедбіадаыедр QR aH анато opu pe iile dili ii Fig. Fig. Fig. OMPHALOCARPUM, AND ASTERANTHOS. 21 24. Тһе radicle and plumule removed : ай nat. size. 25. А seed of Napoleona Heudelotii, showing its different form. 26. One of its cotyledons, with the immersed radicle: nat. size. Both copied from Jussieu's drawing. Section B. Analysis of Asteranthos brasiliensis. 1. A portion of a branchlet, showing a leaf with its axillary flower. 2. A flower in bud, showing the mode of estivation of its plicated corolla. 3. The same, when full-grown and naturally expanded, showing the three series of many tree stamens seated round the disk on the base of the corolla, as seen from above: all nat. size. 4. Five of the stamens removed: all nat. size. 5. The same: magnified. 6. Тһе calyx, disk, superior 8-celled ovary, style, and stigma : nat. size. 7. The same, seen in a longitudinal section, to show the ovules suspended from the placenta in the axis of each cell : magnified. 8. The same, as seen from above, showing the disk round the deeply 8-grooved conical ovary: doth magnified. 9. The cup-shaped, 8-toothed, hollow stigma, seen also in longitudinal section: more highly magnified. 10. An ovule: much magnified. Prate IV. Analysis of Omphalocarpum elatum. l. A portion of the plant. 9. A fascicle of flowers, taken from the trunk of the tree. 3. The calyx. | 4. The pseudo-tubular male corolla, showing the 5 petals agglutinated by their long claws to a mem- branaceous tubular disk. | 5. Тһе same cut open, showing 5 phalanges, each consisting of 6 equal stamens, seated on the margin of the agglutinated disk opposite each petal, with an intermediate fringed staminode alternating between each series. 6. A portion of the disk supporting its 6 fertile stamens and 2 of the staminodes : all nat. size. 7. Three of the same stamens : magnified. 8. Transverse section of an anther before and after dehiscence: more magnified. 9. The abortive ovary and style: nat. size. 10. The same: magnified. 11. The corolla of the female flower, similar to that of the male in every respect except that the stamens are all anantherous. 12. A portion of the agglutinated disk supporting a set of sterile filaments and two of the alternating fimbriated scales. 13. A fertile superior ovary, continuous with a long conical thick style. 14. A longitudinal section of the same, showing the hollow style bearing the channels leading to the many cells of the ovary, and surmounted by the — many-toothed stigma: all nat. size. 15. Shows a magnified view of fig. 13. 16. A transverse section of the ovary, showing its 25 cells radiating round its hollow axis: also - magnified. 22 Fig. 17. 18.” 19; 20. 21. 22. 23. MR. J. MIERS ON NAPOLEONA, OMPHALOCARPUM, AND ASTERANTHOS. А fruit, the smaller specimen in the Kew Museum, viewed sideways, with a quarter segment cut away, to show the conglomerate-like structure of the thick pericarp, the persistent style, the singular hollow axis, and the solitary seed in each cell, attached to the axis. А moiety of the same fruit, seen in transverse section, showing the many radiating cells placed round the large hollow axis, each with a single seed attached in the axes. One of the seeds (like that of Lucuma) shown on its face. The same, shown on its ventral edge, which is grooved where the ascending raphe is imbedded, and in the bottom of which a very small hilum is seen. A longitudinal section of the same, parallel with its faces, showing the enclosed nucleus. The nucleus extracted, seen on its edge. A longitudinal section of the same parallel to the faces, showing the embryo enclosed in thin corneous albumen. . Тһе embryo extracted, seen on its face. . The same, shown on its edge. . À seed from the larger specimen of the fruit in the Kew Museum, shown on its face. . The same shown on its ventral edge, which is channelled by the raphigerous groove, in the bottom of which is seen the hilum, A, and near its summit the perforation, f, through which the chord of the raphe passes into the interior. Near the basal point the micropyle, т, is dis- tinctly seen; this structure is precisely that of Lucuma and Imbricaria, shown by Gaertner. АВ. 1. T + TRANS. Linn. Soc. 2. SER. Вот. $729242700 Mg^ N | i 6 92771490 An сопа. ас иге ЕМ Naciu М J. Miers, del. Lith. London. T a OLEONA. LP A E N с TRANS. LINN. Soc. 2. SER. Вот. ТАв7. NAPOLEONA. TRANS. LINN. Soc. 2.SER. Вот. Tas. 3. Мше k Macdonald, Lath London 22 di Ф : : 2 E (a) т x 5 Е ea 5 2 E © Е m 5 = à Е ee romero > | uU © Р TRANS. Linn. Soc. 2. Sen. Вот. Тав. & % pA 2. -— OMPHALOCARPUM ELATUM. П. On the Алхеттпеге, a new Tribe of the Cordiaceze. By Joun Miers, F.R.S., V.P.L.S., $c. (Plates V.-VIII.) Read June 18th, 1874. THIS memoir, written five years ago, and alluded to on a former occasion *, describes a small group of plants marked by very peculiar characters. They are distinguished chiefly by the unusual growth of the calyx, which often increases in an extraordinary degree as the fruit advances to maturity; but they are more especially remarkable for the atropous development of their ovules and seeds. The group, consisting of the genera Auxemma, Sacellium, Hymenesthes, Patagonula, Paradigma, and Plethostaphia, may thus be characterized : —- AUXEMMEA. Tribus Совртлсвлком. Flores hermaphroditi. Calyx tubulosus, dentatus aut divisus, demum amplifi- catus. Corolla tubulosa. Stamina inclusa. Ovarium superum, disco carnoso hypogyno insitum, 4-loculare : ovula in quoque loculo solitaria, erecta, et atropa. Drupa carnosa, vel subexsucca, mono- pyrena: pyren ossea, rugosa, 4-locularis, aut abortu 2-locularis: semen in loculo solitarium, struc- tura Cordie, sed omnino atropum. l. AUXEMMA. The type of this genus is a curious Brazilian plant found by Gardner, who regarded it as a species of Patagonula, and mentioned it in his * Travels,’ p. 159. Тһе flower, in structure, is not unlike that of the genus just mentioned, has a more tubular calyx, which becomes vastly inflated around the fruit, increasing to about 30 times its original length, and swelling proportionally. The drupe, which it conceals, is nearly of its length, and about the size and shape of a small damson, containing a hard scrobiculated nut, which is 4-сеПей ; its seeds distinctly atropous. The generic name is derived from avfw, amplifico, čupa, vestis, from the singular enlargement of the calyx. AUXEMMA, nob. Сайт parvus, tubulosus, membranaceus, ore 5-dentatus, persistens et in fructu valde amplificatus. . Corolla tubulosa ; tubus cylindricus, calyce dimidio longior, subplicatus, fauce vix ampliore ; /imbi lobi 5 zequales, rotundato-ovati, tubum subzequantes, subexpansi, sestivatione introflexi, cum apicibus incur- vatis centro conniventes. Stamina 5, lobis alterna, inclysa, inequalia, tubo dimidio breviora; fila- menta tenuia, paulo supra basin tubi inserta: anthere ovate, 2-lobe, lobis filamento adnatis, rima laterali utrinque dehiscentibus. Discus hypogynus, carnosus. Ovarium superum, disco insidens, conico-ovatum, 4-loculare; ovula in loculis solitaria, e basi erecta, atropa: stylus brevis, ultra medium 2-fidus, ramis divaricatis bifidis: stigmata 4, oblique clavata, corolla inclusa. Calyz in * Contrib. Bot. ii. 261. 24, МВ. J. MIERS ON THE АОХЕММЕЛ. fructu magnopere auctus, inflatus, reticulato-membranaceus, vesiciformis, plicato-pentagonus, imo ad angulos in saccis prolatus et hinc profunde cordatus, fauce dentibus 5 angulis continuis sensim valde constrictus: drupa ter brevior inclusa, ovata, nitida, sarcocarpio amplo spongiose carnoso vestita, apice subumbilicata et stylo rostrata, monopyrena : pyren osseus, ovoideus, apiculatus, extus echinato-rugosus, apice 4-locularis, imo in cruribus 4 loculis oppositis prolatus, interstitiis cum centro excavato materia medullari farctus: semina in quoque loculo solitaria, atropa, erecta; festa tenui, hilo basilari cum chalaza confuso versus angulum notata; embryo exalbuminosus; radicula supera, brevis, teres ; cotyledones 2, ample, crasso-foliaceze, multiseriatim longitudinaliter plicate. Arbuscula Brasiliensis, frondosa, ramosissima; folia alterna, oblonga, petiolata; panicule in ramulis ter- minales, dichotome ramosissime, multidivise ; flores parvi, plerique тот decidu. 1. AUXEMMA GARDNERIANA, nob.: ramulis subangulatis, striatis, pallide brunneis, gla- bris, axillis subapproximatis: foliis lanceolato-ellipticis, imo sensim cuneatis, apice obtusis, grosse et remote dentato-sinuatis, utrinque glabris, supra subnitidis, subtus pallidioribus, nervis prominulis; petiolo tenui, limbo Әріо breviore: panicula in ramis alaribus laxa, 5-біев dichotome divisa, artubus ultimis tenuissimis scor- pioideis; floribus parvis uniserialiter sessilibus, unico maturescente, reliquis deci- duis; calyce breviter 5-dentato, dense rigide pubescente; corollee tubo membra- naceo, 5-plicato, glabro, fauce ampliore, limbi lobis rotundato-oblongis, expansis, tubum :quantibus, utrinque (sed extus densius) rigide pubescentibus ; staminibus 5, жала ав, inclusis, imo tubi enatis, glabris; ovario oblongo, subconico, parvo, dense piloso; stylo piloso : drupa ovata, subcarnosa, nigra, nitida, calyce valde ampliato abscondita, monopyrena : cæt. ut in char. gen.—In Brasilia, prov. Ceará : v. s. in hb. Mus. Brit. inter Aracati et Ico (Gardner, 1779). This is described by Gardner in his travels (/oc. cit.) as a tree growing у and abundantly in the province of Ceará, where it is known by the name of Pdo branco (white wood) ; it grows to heights of from 20 to 40 feet, its wood, not very hard, being ex- tensively used by the natives for fuel. The branchlets are 2 lines thick, with axils 1-11 in. apart; the leaves are 5-63 in. long, g; panicle is about 3 in. long and broad, 5 or 6 times dichotomously divided; the ultimate divisions are very slender, 3 in. long, bearing on one side about 8 or 10 nodules where the sessile flowers were attached; but in the dried specimen only a single flower remains upon the spikelets; the calyx is 2 lines long, 11 line broad; the tube of the corolla 2 lines long, the lobes of the border 2 lines long, 1 line broad; the filaments 3 line, the ovary $ line long, the style of the same length. In fruit the enlarged calyx is 21 in. long, 1$ in. broad, membranaceous and reticulated, with 5 acute angles, terminating below in 5 large hollow sacs which extend half an inch below the summit of the thick- ened spikelet, tapering upwards to a width of 4 lines, and terminated by 5 erect teeth; the drupe, fixed on the point of atjachment, is 2 in. long, 8 lines broad; the sarcocarp thick, soft, and spongious, enclosing an osseous oe with an uneven surface, with a hollow space below, filled with medullary matter; above, it has four cells, two of which are smaller, but all four contain each a single erect seed ; in the base of each cell a foramen is ‚ seen for the passage of the nourishing vessels to the basal chalaza of the seed, which is nearly 3 lines long ; its two integuments are white, very thin, and brittle, marked at the pointed summit by the mieropyle; there is no trace of a raphe or other spiral vessels. а MR. J. MIERS ON THE AUXEMMEZ. bo сл 2. SACELLIUM. This genus was established in 1808, by Bonpland, in the * Plantee Hquinoctiales,’ where there is an imperfect description and a drawing of the typical plant. From the few particulars known of it, there can be no doubt that the genus closely approximates to Auxemma, agreeing with it in its enlarged inflated calyx enclosing a small monopyrenous drupe, with a 4-celled nut of similar structure. On account of the extreme difference in the appearance of the flower in its early and later stages, Bonpland erroneously concluded that the inflorescence was monecious, bearing male and female flowers in the same panicle; but his evidence does not support this conclusion. Bonpland, who has no high claim as an observing botanist, confesses that such details of the floral structure he owes to Humboldt, who made the drawing above referred to. Kunth, 17 years later, examined Humboldt’s specimen, which had no flowers, and he had only the original sketch of the latter to guide him. In regard to the fruit, he ascertained that its nut had 4 seminiferous cells in its summit ; and he noted the existenee of the inferior cavities figured by Humboldt, in which he found no trace of seeds. Judging by analogy, we have here, therefore, precisely the same structure as that I have described in Ausemma. Upon these materials а reformed generic character is constructed in the following manner :— SACELLIUM, Humb. & Bonpl. Flores hermaphroditi. Calyx primum parvus, tubulosus, 5-dentatus, persistens et in fructu magnopere augescens. Corolla brevissime tubulosa, fere ad basin 5-fida, laciniis oblongis. Stamina 5, imo tubi affixa, brevia, subinclusa. Ovarium superum, disco hypogyno crassiusculo insitum, 4-loculare, loculis l-ovulatis; stylus bis bifidus; stigmata 4, parva. Calyx fructifer maxime auctus, oblongus, vesiceformis, submembranaceus, 5-plicatus, angulis nervosis in dentibus 5 parvis os stringentibus terminatus: drupa quater brevior, inclusa, oblonga, nitida, disco stipitata, stylo longe apiculata, monopyrena; pyren ovatus, rostratus, osseus, summo 4-locularis, imo in cruribus 4 protensus, inter- stitiis et centro cavato medulla repletis, loculos alteros fingentibus. Semina in quoque loculo solitaria, conice ovata, erecta: integumenta alba, tenuissima ; embryo exalbuminosus, atropus; radicula parva supera, obtusa; cotyledones 2, magne, pluriseriatim longitudinaliter plicate. Arbuscula Peruviana, frondosa : folia alterna, lanceolata, petiolata: panicula supraazillaris, dichotome ramosa; flores minimi, ^ 1. SACELLIUM LANCEOLATUM, Humb. & Вопр. РІ. Aquin. i. 47, tab. 18; Lam. Dict. Suppl. v. 10, tab. 995; H., B. & К. vii. 207; DC. Prodr. ix. 502: arbuscula ramosissima, foliis lanceolatis, utrinque acutis, summum versus remote dentatis, submembranaceis, nervoso-retieulatis, supra hirtellis, subtus pallidioribus, molliter pubescentibus, costa nervisque prominulis; petiolo tereti, canaliculato, pubescente, limbo 15-18р10 bre- viore : panicula supraaxillari, folio ter breviore, dichotome ramosa, pubescente ; floribus parvis, subcrebris, brevissime pedicellatis; calyce fructifero valde aucto, acute ovato, reticulato-nervoso, membranaceo, pentagono, 5-plicato: drupa 6plo breviore ovata, rostrata: eset. ut in char. gen. In Peruvia alta, ad uid Те bamba prope охаш (non vidi). | This is described as a tree 9-12 feet high, with a copiously frondose taka: and a trunk with white wood, 1 foot in diameter. The axils are about 2 in. apart; the leaves are SECOND SERIES.— BOTANY, VOL. I. E | . 26 MR. J. MIERS ON THE AUXEMME X: about 6 in. long, 14-15 lines broad, on petioles 4-5 lines long; the flowing panicle issues at a point 1 or 2 lines above the axil, is 21 in. long; thecalyx is 1 line long, the corolla very little longer; the fructiferous calyx is membranaceous, reticulated, of a straw- colour, about 1 inch long, 7 lines broad, encloses a drupe 3 lines long, of a blackish brown colour, and polished. 3. HYMENESTHES. This genus is proposed for a plant collected in Cuba by Wright, and named “ Bour- гена succulenta, Jacq." by Dr. Grisebach, who regarded it as the flowering specimen of another plant to which he gave the same name, both having been distributed, with tickets thus denominated, under Wright's number 3119, as they appear in the herbaria of the British Museum and that of Kew. Тһе latter plant is certainly named in error, and has been described by me as Bourreria clariuscula *; the former plant does not even belong to that genus t, nor to the same family, thus adding another to the many misnomers already specified in Dr. Grisebach's enumeration of Wright's Cuban plants. "The struc- ture of the flower in the plant in question differs extremely from that of Bourreria, and approaches that of Patagonula. Тһе flowers are small; the calyx and corolla are of extremely thin texture and transparent; the tube of the latter is as long as the calyx, its border, of the same length, is divided into 5 linear-oblong segments, which are deeply introflexed and connivent over the stamens; the exserted anthers stand upon equal subulate filaments, pilose below, and inserted near the base of the tube; the ovary, style, and stigma as in the preceding genera. As all the flowers are in a young state, we have no indieation of the growth of the calyx or the strueture of the fruit. The genus differs from Patagonula in the longer tube of the corolla, in the variable number of its parts, and other particulars. Its name is derived from оит», membrana, £a0nc, vestis. HYMENESTHES, nob. Calyx campanulatus, textura tenui, ore 3- vel 4-dentatus, dentibus rotundatis. Corolla tubulosa, tenuis- sime membranacea, tubo turbinato, calyci zequilonga, limbi lobis 4-5-6, oblongo-acuminatis, margi- nibus crenulato-sinuatis, valde reflexis, in :estivatione profunde introflexis et centro conniventibus. Stamina numero loborum, iis alterna, exserta ; filamenta subulato-filiformia, imo compressa, patentim hirsuta, superne glabra, infra medium tubi inserta; anthere ovate, 2-lobe, utrinque rima laterali dehiscentes. Ovarium conicum, disco magno cylindrico continuum, 4-loculare ; ovula in loculis solitaria, e basi erecta, atropa. Stylus subbrevis, fere ad basin bifidus, ramis paullo divergentibus, iterum fissis; stigmata 4, parva, compressa. Cet. ignota. Suffrutex Cubensis, glaberrimus : folia alterna, obovata, nitida, petiolata: panicula terminalis, dichotome divisa; flores parvi. 1. HYMENESTHES NITIDA, nob.: Воштгегіа succulenta in parte, Griseb. (non Jacq.), . Cat. Pl. Cub. 209: ramulis teretibus, subrugulosis; foliis obovatis, imo ad medium sensim attenuatis, apice rotundatis aut emarginatis, crassiusculis, utrinque glaber- rimis, supra nitidis, lete viridibus, valde reticulatis, subtus pallidioribus, nervis venisque prominulis ; petiolo subtenui, limbo 6plo breviore: paniculis terminalibus 222% Contrib. Bot. ii. 285. | | t Ibid. р. 233. MR. J. MIERS ON THE AUXEMME2. 27 dichotome multidivisis, ramis tenuibus, glabris: floribus parvis, cum pedicellis bre- vissimis articulatis ; calyce 3-dentato; corolla tenuissime membranacea, limbi lobis et staminibus inconstantibus, extus obsolete puberula. In Antillis: v. s. in hb. Mus. Brit. et Hook. ins. Cuba (Wright 3119, sub nom. Bourreria succulenta). The leaves are 13-21 in. long, 1-1} in. broad, on petioles 3—4 lines long; the panicle is 15 in. long; the calyx is 2 lines long, 13 line broad ; the tube of the very transparent corolla, as well as its lobes, are of the same length ; the lobes in the flowers of the same panicle, as well as the stamens, vary in number from 4 to 6, the rounded teeth of the calyx being constantly 3. 4. PATAGONULA. The typical species of this genus was first made known by Dr. Dillen in 1732, who described it from a cultivated plant raised in Dr. Sherard's garden at Eltham, and figured by him in his Hort. Elthamensis. After the fashion of that period, he called it simply Patagonica, because he learned it had been introduced from Patagonia. No original specimen of the plant has ever been obtained, so that its exact locality is not known. 1t must be remembered that the date of Dillen's plant is 36 years prior to Captain Cook's first voyage with Dr. Solander, who was the first regular botanist to explore the Tierra del Fuego and the Straight of Magellan. We must therefore look back to the days of Dampier, who visited the coasts of Patagonia several times between 1700 and 1720, and who, it is known, had a taste for botany, and collected plants and seeds in the course of his numerous voyages. At that time, and for nearly a century afterwards, the shores of the whole of the South American continent remained utterly inaccessible to European search ; and it is, therefore, with the greatest probability that we may assign the original source of the plant cultivated in Sherard's garden to the seeds brought home by Dampier from Patagonia, perhaps Tierra del Fuego. We know that a kindred species was obtained in 1820 by Sello, in South Brazil, and described in 1829 by Chamisso; but that is a far more glabrous plant, with narrower leaves, and is certainly very distinct. Patagonula, as a genus, was first established by Linneus in 1787, who gave it that substantive name in lieu of Dillen's adjective prefix. It was also enumerated in Aiton's * Hortus Kewensis’ in 1789, under the name of Cordia Patagonula. Jussieu, in 1789, placed Patagonula near Cordia; Endlicher (1838) considered it a doubtful genus: and DeCandolle (in 1856) excluded it from the family without assigning to it another position. Prof. A. DeCandolle was of the same opinion, because he had observed its erect ovules in Blanchet's specimen from Bahia, on which account he was disposed to refer the genus to the Verbenacem. This opinion was not supported by Schauer, who soon Mai composed his monograph of the latter family for the Prodromus. Fre- senius, in 1857, in the * Flora Brasiliensis, placed Patagonula in the Cordiacee without hesitation, though ‘he does not seem to have examined it. It appears that no one has taken the trouble to inquire into the subject ; and I therefore give the following diagnosis, founded upon my own analysis. E 2 „кое былды йй 28 MR. J. MIERS ON THE АОХЕММЕ Ж. | PaTAGONULA, Linn.; Patagonica, Dill. Calyx parvus, fere ad basin 5-partitus, laciniis zequalibus, demum in fructu valde auctis. Corolla tubu- | losa, tubo calycem zquante subcampanulato, subplicatim 5-sulcato, limbi laciniis 5, ovatis, obtusis, ] expansis, in wstivatione apicibus incurvis centro conniventibus. Stamina 5, laciniis alterna ; fila- menta filiformia, imo latiora, supra basin tubi inserta, exserta; anthere ovate, 2-lobe, introrse, medio dorsi affix, longitudinaliter dehiscentes. Ovarium superum, conice ovatum, disco 4-lobo | insitum, 4-loculare, ovulis in quoque loculo solitariis, e basi erectis, atropis. Stylus bis bifidus ; | stigmata 4, parva, subelavata. Drupa parva, exsucca, ovato-globosa, stylo apiculata, calyce valde aucto subcoriaceo nitido stellatim dilatato cireumdata, monopyrena ; pyren subglobosus, imo breviter stipitatus, apice rostratus, osseus, 4-locularis, loculis 2 зере abortivis: semen conico-oblongum, apice micropylo notatum, ad basin angulum versus hilo cum chalaza subconfuso affixum ; integu- menta 2, tenuia, albida, raphe nulla; embryo exalbuminosus, radicula tereti, supera, cotyledonibus 2 longitudinaliter multiplicatis multo breviore. Arbuscule Americe meridionalis, ramosissime : folia elliptica vel lanceolato-oblonga, glabra aut pilosula, sepe dentata, petiolata : panicule azillares vel terminales, dichotome multidivise, sepe pubescentes, ramis tenuissimis ; flores parvi, ebracteati. 1. PATAGONULA AMERICANA, Linn. Gen. Pl. (1737) р. 45; Poir. Dict. v. 57; Lam. Illust.. 424. n. 1915, tab. 96; Cham. in Linn. iv. 492; Fres. in Mart. Fl. Br. fasc. xix. p. 27: Patagonica, Dill. Elth. (1732), p. 304, tab. 226. f. 203: Cordia Patagonula, Dryand. in Ait. Hort. Kew. 1. 259: Petrea dentata, Spr. Sys. ii. p. 76: ramosissima, ramis tere- tibus, tenuibus, striatis, lenticellis albidis rugulosis, patentim longe rufo-hirtulis, rarius demum subglabris, in axillis presertim hirtis; foliis alternis, interdum binis, oblongo-ellipticis, aut lanceolato-oblongis, imo angustioribus, apice obtusis vel sub- acutis, seepius remote grosse serratis, marginibus nervigeris plus minusve ciliatis, submembranaceis, valde reticulatis, utrinque glabris, supra pallide viridibus, in costa nervisque suleatis, subtus pallidioribus, costa nervisque prominulis, in junioribus minute pellucido-punctatis ; petiolo tenui, canaliculato, limbo 15-20plo breviore: paniculis axillaribus et terminalibus, dichotome multidivisis, ramis tenuibus, rugu- losis, hirsutis; floribus parvis, pedicellatis; calyce pallide membranaceo, fere ad basin 5-fido, lobis oblongis, erectis, extus glandulis fuscis piliferis punctatis, margine longe ciliatis; corolla glabra, tubo campanulato, calycem sequante, laciniis 5 :equi- longis, acute ovatis, expansis, minute pellucido-punctatis; staminibus imo insertis, exsertis; ovario glabro, subconico, 4-loculari; stylo bis bifido: drupa parva, api- culata, calyce valde aucto cireumdata, laciniis spathulato-oblongis, emarginatis, reti- culatis, radiatim expansis. In Patagonia: v. s. in hb. Mus. Brit. ex hb. Mill. in hort. Warner. cult. This species is well distinguished from P. glabra, which has been confounded with it, by its larger, more membranaceous, more deeply dentate leaves, hispid branchlets, and many other characters. The axils of its branches are 1—3 in. apart; the leaves are 2-21 in. long, 3-15 in. broad, on slender petioles 2 lines long. The panicles are 1-13 in. long; the pedicels 3 line long; the calyx 1 Нпе long; the tube of the corolla and seg- ments each 1 line; the lobes of the fructiferous calyx, according to Dillen, are 6 lines long, 2 lines broad. . MR. J. MIERS ON THE АСХЕММЕЖ. 29 2. PATAGONULA GLABRA, nob. : Patagonula Americana, var. glabra, Cham. Linn. iv. 492 ; Fresen. in Mart. Fl. Br. l. c. p. 27, tab. 9. fig. 9 : ramosissima, ramis tenuibus, striolatis, lenticellato-rugulosis, glabris: foliis lanceolato-oblongis, imo subeuneatis, apice obtusis aut subacutis, junioribus lineari-lanceolatis, summum versus remote dentatis, marginibus revolutis, subcoriaceis, impunctatis, opacis, utrinque glaberrimis, supra viridibus, subtus pallide glaucis, costa nervisque prominulis, venis tenuissime reticulatis, interstitiis elevatis, hine quasi punctato-granulatis ; petiolo tenui, supra plano, limbo 12-15plo breviore: paniculis axillaribus et terminalibus dichotome multidivisis, glabris, raro obsolete pilosulis; floribus parvis, pedicellatis ; calyce 5-partito, lobis oblongis, obtusis, extus pilis articulatis puberulis, apice pilis sim- plicibus ciliatis; corolle tubo calycem :equante, campanulato, extus glabro, intus scabridule pruinoso, limbi lobis 5, ovatis, tubo æquilongis, expansis; staminibus 5, in tubi basin insertis, ultra faucem vix exsertis, glabris; ovario conico, disco 4-lobo imposito, 4-loculari; stylo bis bifido: drupa parva, conice globosa, imo stipitata, apice apiculata, calyce demum magnopere aucto circumdata, laciniis spathulato- oblongis vel lineari-oblongis, reticulatis, pellucido-punctulatis, glaberrimis, radiatim expansis. Іп Brasilia, prov. Rio Grande do Sul: v. s. in herb. Mus. Brit. et Hook. Encrucilhado, Rio 8. Barbara (Sellow). Sellow's plant was considered by Chamisso a mere variety of the typical species, which he had never seen; but it differs widely in its much smaller, more coriaceous leaves and other characters. It forms a tree of moderate size, furnishing a useful timber of a yellowish white colour, called Zpé branco. Its slender branchlets are quite glabrous, a few hairs being sometimes seen on the axillary bud; the axils are 1—4 in. apart; the leaves, which Chamisso compares to those of the Laurel, are 11-21 in. long, 7-12 lines broad, on petioles 13—2 lines long: from their opaque texture they form a great contrast to the submembranaceous leaves of the preceding species. The panicles, which have slender ramifications, are almost glabrous, about 2 in. long; the calyx is said by Fresenius to be covered with glanduliferous hairs. I found the corolla persistent at the base of the free drupe; it has a campanular tube 1 line long, with lobes of equal length; the filaments are not fixed in the throat, as Chamisso states, but near ‘the base of the tube. The fructiferous calyx has 5 spathulate lobes 10-11 lines long, 13-2 lines broad, with a central stipitate drupe 2 lines long and 1 line broad. © 3. PATAGONULA TWEEDIANA, nob.: ramulis tenuibus, angulato-striatis, glabris: foliis glaberrimis, lanceolatis vel lanceolato-oblongis, imo cuneatis et integris, a medio sursum gradatim acutis et crenulato-dentatis, chartaceis, supra viridibus, nitidis, costa flavida complanata, tenuiter nervosis, granulato-punetatis, subtus pallidioribus, nervis tenuissimis immersis, creberrime flavido-maculatis ; petiolo tenui, subplano, limbo 12-15plo breviore: paniculis axillaribus diehotome divisis, floribus parvis in ultimis sectionibus glomerulatis; calyce 4-partito; corolle glabrsz tubo brevi sub- campanulato, limbi laciniis 4, oblongis, obtusis, extus obsolete puberulis; fila- mentis tenuissimis 4, paullo sub faucem insertis, longe exsertis, sparse pilosulis : 30 MR. J. MIERS ON THE АОХЕММЕХ. drupa conice globosa, parva, imo brevissime stipitata et bifossata, apice rostrata, corolla persistente cincta et calyce valde aucto 4-lobo cireumdata, 4-loculari (loculis 2 subabortivis) ; seminibus generis. In Brasilia prov. Rio Grande do Sul: v. s. i» hb. Hook. Viamaó ( Tweedie, 27). Tweedie describes this as a “ fine tree," growing in the woods of Viamaó, about 10 miles E. of Port Alegre, and 30 miles from the locality of the preceding species. It is distinguished from the latter by its more lanceolate leaves, by its glomerulate flowers, which are constantly 4-merous, by the insertion of its stamens, its hairy filaments, and the presence of 4 augmented calycine lobes around the fruit. Тһе branches are slender, with axils 4 in. apart; the leaves are 23-3 in. long, 8-10 lines broad, on petioles 23 lines long; the tube of the corolla is 1 line long, its lobes of the same length; the 4 patent segments of the augmented calyx are 10 lines long, 25-3 lines broad; the exsuccous drupe is 3 lines long, 2 lines broad, is shortly stipitate at the base, with 2 hollow cavities; the nut, somewhat fibrous in texture, has 2 larger cells, bearing each an erect seed, terminated by the micropyle, and fixed by its base to the bottom angle of the cell, with a basal ehalaza, no raphe being visible. 4. PATAGONULA BAHIENSIS, Moric. Pl. Amer. tab. 86; A. DC. Prodr. ix. 467, in not.; Fresen. in Fl. Br. /. c. p. 28: ramulis fuscis, lenticellis albis rugulosis, glanduloso- tomentosis: foliis alternis, vel 2-3 approximatis, ovato-ellipticis, utrinque sub- acutulis, imo circa petiolum brevissimum subito rotundiusculis, acumine obtusiusculo, fere integris, aut in marginibus cartilagineis crenulatis obsolete denticulatis, supra sublucidis, fere glabris, costa nervisque subimmersis hirtellis, subtus pallidis, opacis, dense puberulis atque setaceo-hirsutulis, costa nervisque prominulis ; petiolo . semitereti, setaceo-piloso, limbo 40plo breviore : paniculis axillaribus et terminalibus, tenuiter dichotome multidivisis, pubescentibus; floribus parvis, subsessilibus ; calyce 5-partito, lobis oblongis, acutis, extus glanduloso-setaceo hirsutulis, margine ciliatis; corolla membranacea, minute pellucido-punctata, tubo subcampanulato, calycem æquante, limbi lobis 5, ovato-ellipticis, obtusis, subreflexis, margine ciliatis; staminibus 5, imo insertis, vix exsertis, glabris; ovario semigloboso, disco insito, 4-loculari; stylo bis bifido. In Brasilia, prov. Bahia: v. s. in hb. Hook. Jacobina (Blanchet, 3086). | Besides the above cited, DeCandolle quotes others of Blanchet's numbers (2855 and 3902) among his specimens. The axils of its branchlets are 1-11 in. apart; the leaves 8-3% in. long, 1-12 in. broad, on petioles 1 line long; the panicle is 28 in. long and broad; several flowers are agglomerated at the end of its branchlets into a head 3 lines in diam.; the flower expanded is 2 lines broad, the calycine lobes 3 line long; these | : expand in the fruit to а length of 9 lines and breadth of 2 lines, nerved and reticulated ; .the fruit as in the preceding species. 5. PARADIGMA. . This genus is proposed for a plant from the island of Cuba, insufficiently described many years ago by Richard as a species of Cordia. The general structure of the flower and fruit, the peculiar estivation of the corolla, its numerous lobes, its sessile linear ымын аа Бк даны а д. МБ. J. MIERS ON THE АГХЕММЕЖ. 31 anthers, its stigmata, the position of its ovules, and the enlargement of the calyx around the fruit, all prove it to be very different from Cordia. Itis evidently allied to Auxemma, differing in its much larger corolla, in its more coriaceous calycine envelope of the fruit, the structure of which is similar to that of the genus last mentioned. The name given to it is derived from тарадегуна, exemplar, as it offers а striking example of the peculiar characters which distinguish Auzemma. PARADIGMA, nob. Flores hermaphroditi. Calyx ovato-tubulosus, coriaceus, primum omnino clausus, tunc im dentibus brevibus ssepius 3, rotundatis irregularibus margine dense ciliatis apertus, demum accrescens et per- sistens. Corolla majuscula, tubulosa, tubo carnosulo, pro dimidia parte inferiore anguste cylindrico et calycem æquante, seorsum subito infundibuliformi, fauce campanulata, crebre parallele nervosa, limbo 8-lobo, lobis oblongis, acutis, demum expansis, zstivatione tubo multiplicato lobis profunde introflexis et in centro conniventibus. Stamina 8, lobis alterna, sub faucem inclusa ; filamenta nulla ; anthere æquales, lineares, intra faucem sessiles, tubo ter breviores, 2-lobæ, lobis connectivo angusto membranaceo supra medium nexis, imo segregatis et ad sinum affixis, utrinque lateraliter dehiscenti- bus; pollen globosum, reticulatum. | Ovarium superum, oblongum, disco parvo insitum, 4-loculare ; ovula in loculis solitaria, erecta, atropa ; stylus filiformis, bis bifidus; stigmata 4, inclusa, majuscula, lineari-lamelliformia, marginibus anguste membranaceis et crenulato-undulatis. Drupa majuscula, exsucca, oblonga, apiculata, calyce coriaceo aucto ore dentato occlusa, monopyrena ; pyren ovatus, apice conico-acutus, lateribus sulcis irregularibus et interruptis sarcocarpio repletis profunde exsculptus, dense osseus, cavitate centrali ampla excavatus, et hine materia fibrosa vasis repleta е disco enata munitus, superne 4-locularis, loculis monospermis (uno alterove minore) ad basin pro vasorum introitu perforatis. Semen loculum implens, oblongo-ovatum, ad basin prope ab angulo affixum : integumenta 2, externum album, friabile, internum crassiusculum, opacum; raphe nulla; chalaza cum hilo basali confusa; embryo exalbuminosus, atropus ; radicula supera, parva ; cotyledones 2, carnosule, plurifariam longitudinaliter plicatze. Arbor Antillana, ramosissima : folia alterna, ovata, integra, petiolata: panicule terminales, dichotome divise ; flores majusculi, speciosi. 1. PARADIGMA GALEOTTIANA, nob.: Cordia Galeottiana, A. Rich. in Flor. Cub. Sagr. хі. 109; Griseb. Pl. Cub. 208; Walp. Ann. v. 540: ramis glabris, substriatis: foliis glaberrimis, ovalibus vel oblongis, apice breviter constrictis obtusis et canalieulatis, | imo obtusis, coriaceis, integris, supra opace viridibus, costa nervisque tenuibus sub- immersis, subtus pallidioribus, nervis patentibus rubellis paullo prominentibus ; petiolo supra sulcato, limbo 6plo breviore : paniculis terminalibus, dichotome multi- ramosis, glabris, ramulis compressis, ultimis 1-2-floris, bracteis lineari-acutis valde deciduis; calyce tubuloso, coriaceo, glabro, intus parallele multinervoso, apice primum clauso, dein aperto, dentibus 3-6 rotundatis, ciliatis, extus obsolete, intus densius pilosis; corolla infundibuliformi, tubo calyce 2plo longiore, imo angustato, glabro, superne ad nervos 24 prominulos pilosulos plicato, limbo expanso, lobis 5, lanceolato-oblongis, acutis, submembranaceis, extus dense cano-tomentosis ; antheris intra faucem affixis, 8, linearibus, sessilibus : ovario subulato-oblongo, striato : drupa oblonga, apiculata, calyce aucto ore contracto tecta, exsucca; pyrene 4-loculari, - 4-spermo. Іп Antillis: v. s. in hb. Mus. Brit. Cuba circa Havanna (Galeotti, 19). 32 MR. J. MIERS ON THE АОХЕММЕЖ. The axils of this plant are about 2 т. apart; the leaves are 21-4 in. long, 14-1 in. broad, on petioles 5-8 lines long, with about 10 pairs of slender patently divaricate nerves; the panicle is 2 in. long; the calyx in flower 5 lines long, 3 lines broad; the tube of the corolla is 14 in. long, 11 line broad at the base for a length of 5 lines, thence gradually widening to a breadth of 9 lines at the mouth; the lobes are 6 lines long; the anthers 3 lines long, 1 line broad; the conical ovary is 2 lines long, $ line hibáik the subulate style 3 lines long, its branches 2 lines; the stigmata 2 lines long; the enlarged calyx concealing the fruit is 9 lines long, 7 lines broad ; the oval drupe, attenuated at its summit, is 9 lines long, 6 lines broad; nut deeply furrowed, has a deep hollow along its axis, and has 4 basal cavities at its base filled with pithy matter, as in Auzsemma. 6. PLETHOSTEPHIA. This genus is proposed for another of Richard’s Cuba plants, his Cordia angiosperma, which differs from that genus in its large salver-shaped corolla, with a rotate border of 15 lobes, with their margins overlapping one another dextrorsely, in its 10 included stamens in 2 unequal series, its 10-lobed hypogynous disk supporting a 4-celled ovary. The calyx is tubular, and 5-toothed, becoming thicker and enveloping the fruit as in the preceding genus. Ав I have not seen any specimen, the generic character is founded wholly on Richard’s description and figures. Its name is derived from zA56oc, multitudo, зтефе corono, because of the many-lobed border which crowns the corolla. PLETHOSTEPHIA, nob., Со’ а sp., Rich. Calyx tubulosus, margine 5-dentatus, persistens et augescens, Corolla subhypocrateriformis, tubo calyce 3—4plo longiore, imo cylindrico, sursum anguste infundibuliformi, fauce nuda, limbo patentissimo, tubo 5рю breviore, in lacinias 15 diviso, /aciniis recte oblongis, acutis, marginibus (sec. icon.) dextror- sum imbricantibus. Stamina 10, inzqualia, alternatim biseriata, inclusa; filamenta subbrevia, teretia, erecta, supra medium tubi inserta; anthere ovate, filamento ter breviores, 2-lobz, lobis adnatis introrsum dehiscentibus. Discus hypogynus, pulvinatus, crenatim 10-lobus: ovarium superum, oblongum, disco insitum, 4-loculare ; ovu/a in quoque loculo solitaria, erecta, atropa; stylus filiformis, corollz tubo zquilongus, apice 4-fidus, laciniis recurvis ; stigmata 4, parva, capitata. Drupa conice ovata, carnosa, imo stipitata, stylo immutato recurvo apiculata, calyce aucto incrassato occulta, monopyrena; ругеп osseus, acute ovatus, imo disco stipitatus, summo apiculatus, extus sulcis profundis exsculptus, abortu 2-locularis; semen in quoque loculo unicum, basi stipitato-affixum, erectum. Arbor Antillana, ramosa : folia coriacea, late ovata, undulato-crenata, petiolata: paniculæ terminales, dichotome divise ; flores pedicellati, speciosi. l. PLETHOSTEPHIA ANGIOCARPA, nob.: Cordia angiocarpa,, A. Rich. in Fl. Cub. Баст. 119, tab. 60; Walp. Ann. v. 540: ramulis validiusculis hirsutissimis: foliis in apice ramorum subconfertis, late ovalibus aut suborbicularibus, imo obtusis et in petiolum breviter acutatis, apice rotundatis, integris, vel ad margines subundulato- sinuatis, nervis utrinque 6 divergentibus, venis parallele transversis, utrinque hirto-scaberrimis; petiolo late suleato, limbo 7plo breviore: panicula terminali, brevissime peduneulata, quater dichotome divisa, ferrugineo-hirta; floribus spe- * MR. J. MIERS ON THE АОХЕММЕЖ. 33 ciosis, pedicellatis ; calyce tubuloso breviter 4-dentato, pedicello 3plo longiore, ferrugineo-hirto, demum in fructu aucto; сого tubo calyce 3plo longiore, striato, hirto, limbi rotati laciniis 14—15, oblongis, acutis; staminibus 10, omnino inclusis, supra medium tubi insertis; ovario ovato, disco insito; stylo tenui corolle longi- tudine: drupa oblonga, acuta, subcarnosa, imo disco crasso stipitata, apice stylo recurvulo apiculata, calyce aucto incrassato inclusa; pyrene valde exsculpto abortu 2-loculari, loculis monospermis. In Ins. Cuba (пов vidi). Its stout branchlets have their axils 3-4 lines apart; the leaves are 31-5 in. long, 21-3 in. broad, on petioles 9-11 lines long; the panicle is 31 in. long, the peduncle 2 lines long, the ramifications 4 lines long, the ultimate pedicels 2 lines long; the calyx is 5-7 lines long, 3 lines broad; tube of corolla 12 in. long, contracted at its base for the length of 5 lines, where it is 2 lines broad, gradually enlarging towards the mouth, which is 6 lines broad; the horizontally expanded segments of the border are equally 4 lines long. 2 lines broad; the filaments are inserted about 2 lines above the middle of the tube alternately a little above the others; the filaments 21 lines long, the anthers 15 line long; the ovary 2 lines long, on a disk 1 line high; the style 17 lines long, its 4 recurved branches 2-3 lines long ; the enlarged calyx is 12 lines long, encloses a drupe, which, with its discal support, is 10 lines long, 5 lines broad. А drawing of this plant, with its analytical details, copied from Richard's figures, is given in Plate VIII. of the present memoir. . DESCRIPTION OF THE PLATES. Praw V. — Shows a branchlet of A4uzemma Gardneriana, with its dichotomously divided panicle 8 of small flowers. ‚ A flower. . The calyx and corolla, seen separately: all nat. size. . The calyx : magnified. The tubular corolla, with its segments inflected in estivation. . The same, after expansion, when the segments become reflected. The same, cut open to show the position of the stamens. The five unequal stamens, shown separately : all equally magnified. The same: more magnified. The ovary seated on the hypogynous disk, with the style and stigmata. . A longitudinal section of the same. . A transverse section of the 4-celled ovary: all equally magnified. | . The persistent calyx, now grown to 30 times Из, former length, becoming a reticulated bladder- like envelope, deeply 5-angled, narrowed by 5 teeth 1 in the mouth, and NY cordate at the base. 13. The same, with a portion cut away to show the enclosed drupe. - 14. The same, with half of the calyx cut away trausversely, showing its angular form around the enclosed drupe. SECOND SERIES.— BOTANY, VOL. I. Fig. © ©; фе фы — ы tmc F 34 MR. J. MIERS ON THE АОХЕММЕЛ. Fig. 15. The fleshy drupe, shown separately. 16. Half of the pericarp of the same removed, showing the enclosed muricated osseous nut. 17. The nut cleared of all the pericarp, when its four basal fangs are seen round the hollow Базе, 18. A longitudinal section of the same, showing an erect seed in each cell, attached at its base by a broad hilum, into which the nourishing vessels from the basal cavity pass through distinct perforations. 19. A transverse section of the nut, with its 4 cells, each containing а single erect seed. 20. One of the seeds: ай nat. size. 21. The same, with the integument removed, leaving the exalbuminous embryo, consisting of a small superior radicle and 2 very plicated cotyledons. 32. The same, seen from above: both magnified. Puate VI. Section A. Shows a branch of басе ит lanceolatum, in flower and fruit: copied ; from the drawing of Bonpland. Fig. l. A flower: nat. size. 2. The calyx: magnified. 3. Тһе corolla. 4. Тһе same, cut open, to show the position of its 5 equal stamens. 9. The ovary seated on a fleshy disk, surmounted by the style and stigmata: all equally magnified. 6. The much enlarged vesicoid calyx, contracted in the mouth by 5 teeth, and surrounding the drupe fixed at its base. 7. The same, with half of the calyx cut away, showing the drupe placed in the centre, as viewed from above. 8. Тһе drupe removed, apiculated by the style, and seated on the stipitiform disk. 9. A longitudinal section of its enclosed osseous nut, showing in its upper portion 2 of its 4 cells, each with an erect seed : the lower portion is hollow between 4 descending fangs (as in 4uzemma); and it was this hollow part which Bonpland mistook for 4 sterile cells. | 10. A transverse section of the same, across the 4 equal cells. 11. One of the seeds, with a broad basal hilum: all nat. size. 12. The embryo, deprived of Из integument, formed like that of 4uzemma, with a small superior : radicle, and 2 very plicated cotyledons. . 13. Тһе same, as seen from above: both magnified. Section B. А branchlet of Hymenesthes nitida in flower. Fig. 1. A flower: nat. size. 2. The calyx, | 3. The corolla, with its segments inflected in æstivation. 4, The same, after expansion. Ў The same cut open, to show the position of the stamens. 6. The ovary, disk, style, and stigma: all equally magnified. 7. 'The stamens, seen in different positions. 8. The ovary, seated on the disk, surmounted by the style and stigma. i A longitudinal section of the same, showing 2 of its 4 cells. 10. A transverse section of the same, showing its 4 cells : ай more magnified. Fig. MR. J. MIERS ON THE AUXEMMEX. 35 Рглте УП. Section A. shows a branchlet of Patagonula bahiensis with its dichotomously divided panicle. . A flower: nat. size. The same: magnified. . The calyx. . The corolla. The same, cut open to show the position of the stamens. . The ovary upon the disk, with the style and stigma: all magnified on the same scale. . Three of the stamens before and after dehiscence : more magnified. . The fruit, surrounded by the 5 persistent lobes of the calyx, now greatly enlarged and horizon- tally expanded : nat. size. OND op бо фо н Section В. A branch of Patagonula Tweediana, with its dichotomous panicle in fruit. Fig. 1. A persistent flower found at the base of one of the fruits: nat. size. 2. The same magnified, and cut open, to show the position of its 4 stamens, alternate with 4 segments. . The stamens shown separately : also magnified. . A drupe surrounded by the 4 persistent, and now much enlarged, lobes of the calyx. . The dry drupe detached: doth nat. size. . The same, with part of the calycine lobes: magnified. The same detached, and supported by the stipitate disk. . The same viewed from below, to show the two basal hollows. . A longitudinal section of the same showing in each cell an erect seed, fixed at its base, where the nourishing vessels, proceeding from the torus, perforate the nut, to reach the hilum. 10. A transverse section of the same, showing 2 fertile and 2 sterile cells. 11. One of the erect seeds, the chalaza confounded with the basal hilum: all magnified on the same scale. 12. The same enlarged. 13. The same, with its integument removed, exposing the exalbuminous embryo, like that of Auxemma. l4. The same viewed from above: all more magnified. с оз с оо с Prate VIII. Section А. А branch of Paradigma Galleottiana, with Из dichotomous corymb, drawn Fig. from an original specimen collected by Galleotti. l. A flower in bud. 2. The calyx. 3. The corolla beginning to open. 4. The same cut open, showing how the 8 segments аге inflected in estivation, and the relative position of the 8 stamens. The corolla, fully grown and expanded, is seen in the drawing of the plant. 5. The ovary, disk, style, and stigmata: ай nat. size. 6. The same: magnified. 56 MR. J. MIERS ON THE AUXEMME. 7. One of the sessile stamens, formed of 2 linear anther-cells, connected by a shorter intermediate membranaceous connective, attached at its sinus to the corolla : more magnified. 8. One of the 4 compressed linear stigmata, cinctured by an undulating membrane : much magnified. 9. The persistent calyx, now much enlarged, and enclosing the drupe. 10. The fleshy drupe removed. | 11. The fleshy pericarp of the same taken away, leaving а deep foveolated osseous nut. 12. A transverse section of the nut, showing 2 fertile and 2 sterile cells, placed round the hollow axis. 13. А longitudinal section of the same, showing the hollow axis and the 2 fertile cells, each with an erect seed fixed at its base: ай nat. size. 14. A magnified view of the same, showing the nourishing vessels, ascending from the torus, and perforating the nut, to reach the chalazal point of attachment of each seed. 15. The erect seed, separated. 2 : 16. The same, deprived of its integuments, leaving an exalbuminous embryo, like that of Auxemma: both magnified. Section B. Portion of a branch of Plethostephia angiocarpa, with its dichotomous panicle: copied from Richard’s drawing. Fig. 1. A flower. 2. The calyx. 3. Half of the corolla laid open, showing 8 of its 15 segments, and 5 of its 10 stamens, in 2 series. 4. The calyx half cut away, to show the enclosed ovary upon its disk, and surmounted by the style and stigmata: all nat. size. 5. A longitudinal section of the ovary and disk, showing an erect ovule in each cell. 6. A transverse section of the same, showing its 4 cells: both magnified. 7. The persistent calyx, now much increased, and enclosing an oval drupe. 8. The drupe with the fleshy pericarp removed, leaving an osseous foveolated nut, apiculated by the style. 9. A transverse section of the same, showing 2 fertile and 2 abortive cells: all nat. size. 10. A longitudinal section of the same: magnified. a w o. j % 2 5 ay „ 99 : 44 9,2 с err A Macdonald, Lath Londo әм (А b 1,4 7] Maciure I3 5. D 2 AUXEMMA GARDNERIANA. ерләр. ‘порт чит preuopoe % OPN -LANCEOLATUM. HYMENESTHES -NITIDA. 'g'av] Log L Тесеабағав ышы > 7 Bec cL» 3 y Ч Уасімге & Macdonald, Lath London. | > ‘TSP ‘лап f IN т i d У эл > е T asks ат, «Оу E сасна fo PLE THOSTEPHIA ANGIOCAR Я А шрот] решор PARADIGMA GALLEOTTIANA. 9SVL 104 ^39 7 0095 ММГ] SNVH[ THE TRANSACTIONS OF ІНЕ LINNEAN SOCIETY OF LON DO N. SECOND SERIES. BOTANY ; VOLUME I. PART THE SECOND. MISSOURI BOTANICAL GARDEN. LONDON.: PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER- ROW. M.DCCC.LXXV. CONTENTS. PART II.—DECEMBER, 1875. ПІ. On the Origin of the prevailing Systems of Phyllotavis. By the Rev. GEORGE ЗН ЖА FLS По о IV. On the Barringtoniacem. By Joun Miers, F.R.S., F.L.S., Dignit. ей Commend. Ord. Imp. Bris. оне, e: -c na аа 4 ox сз M [37 | ПІ. On the Origin of the prevailing Systems of Phyllotaxis. By the Rev. GEORGE | Hrnstow; M.A., F.L.S., F.G.S. (Plate IX.) Read February 4th, 1875. INTRODUCTION.—In my paper “ On the Variations of the angular divergences of the Leaves of Helianthus tuberosus” (Trans. of the Linn. Soc. vol. xxvi. p. 647), the con- cluding paragraph commenced as follows :—* If we limit our inquiries to the conditions of the primary [ordinary fractional] series alone, we find, as a fact in nature, that if we assume any leaf as No. 1, then No. 2 lies at an angular distance from it, between 120° and 180° inclusively. If we ask why it is so, no answer can be given. Further, starting with this condition, we find the second leaf has an inherent tendency to take up a definite point as its position along that are. ..... If we ask why it is so, or why the second leaves are not anywhere upon that arc, there is as yet no answer to this question any more than to the first.” Іп the present paper I propose endeavouring to reply to both these questions, and to expound what appears to me to be the fundamental origin of the various systems of phyllotaxis exemplified in nature. The Spiral arrangement is deduced from Opposite and Decussate Leaves.—1f a stem (as of the Jerusalem Artichoke) be found having opposite and decussate leaves from the base upwards, but becoming converted into an alternate and spiral arrangement above (a not uncommon feature), the first process of such a change is the development of short internodes, so as to separate the leaves of each opposite pair, as in fig. 1. Тһе numbers represent the leaves, which are supposed to occupy their original decussating positions. | The order of the uwmbers shows the true spiral arrangement now produced by this development of new internodes, and which are indicated by the thin lines, while the original and long internodes are represented by the thicker portions of the spiral. In other words, if a line be drawn from leaf to leaf successively, it will be spiral, and will pass through the leaves in the order indicated by the numbers on the projected helix (fig. 1). (See note, p. 45.) Е In following up this spiral line, it will be seen that just those numbers which are the denominators of the fractions of the usual (primary) series, +, 3, 3, $, 1% зг, fall over one another, viz. 0, 5, 8, 13, 16 [228], and 21. Similarly the number of coils of the helix between 0 and each of these numbers successively supplies the numerators. Thus there are /wo coils between 0 and 5; three coils between 0 and 8; five coils between 0 and 13; and lastly, eight coils between 0 and 21. Тһе denominators of the first two fractions (2 and 3) are not represented amongst these numbers ; the reason for this will be explained further "s In the diagram the leaves are represented as retaining their originally decussating positions; but practically, while the internodes are being developed, the leaves shift their positions, as it were, and obey the following law :— SECOND SERIES.—BOTANY, VOL. I. G 38 THE REV. GEORGE HENSLOW ON THE ORIGIN OF THE The angular distance between any two successive leaves, when projected on the same horizontal plane, is constant.—The rationale of this law seems inexplicable, because we cannot say what causes a leaf to grow out of the axis of a bud at one particular point rather than another. All that can perhaps be suggested is, that as an equable distri- bution of light may be thus presumably acquired by every leaf, so natural selection may have been a determining influence ; but there would seem to be no grounds for assuming external influences to have been the sole originating causes *. Whatever be the ultimate facts which cause the position of leaves to be what it is, they cannot in the present state of our knowledge be determined +. This law is found to be true both for normally alternate leaves and for those which have arisen on a stem by the development of internodes, on which the lower leaves were opposite or whorled in threes. The following observations are the author’s, and quoted from the paper alluded to above (р. 654) :— “ A change from verticils of threes into 3 was frequent. It takes place in the fol- lowing manner :—The first step is to cause the three leaves of the different whorls to separate slightly by a development. of their internodes. Then, if any two consecutive whorls be examined, the order of succession of the six leaves (No. 1 being the lowest leaf) is thus,— i 6 1 in which it will be noticed that the 4th leaf, instead of being over the interval between the 1st and 2nd, is over that between the 1st and 3rd; so that the angle between the * The * Text-Book of Botany,’ by Sachs (translated by А. W. Bennett), p. 179, may be here consulted. t Mr. Hubert Airy has started the proposition that * leaf-arrangement exists for and is determined in the bud." And his “ condensation theory " of leaf-arrangement depends upon “ the contact among the balls which [he] uses to represent embryo leaves.” (* Nature,’ vol. vii. p. 442.) I venture to eall in question both of these fundamental conditions of his theory. That plants generate their buds during a short summer in readiness for the first sufficiently warm weather in spring is an obvious fact ; but that they should be obliged to keep the buds so long undeveloped is no fault of theirs, but a circumstance dependent upon climate. When once a bud is made, and the only opportunity is the сош- paratively short time the plant is in leaf, it matters not whether it remain a week, a month, or six months. The duration сап have nothing to do with the formation of it, or of the importance of any peculiar leaf-arrangement. If, however, we substitute vernation for leaf-arrangement (as synonymous with phyllotaxis), it will be perfectly true; for the half-formed leaves are folded in various ways, so as to to render the bud as compact as possible. Moreover there is nothing in the structure of buds to justify the comparison of embryo leaves to balls. They are not balls at all, but, rather, conical with more or less flat surfaces. Although the apex of a leaf may first appear аза hemisphere, 1 does not retain that form, but elongates into a cone by the time the next leaf emerges from the axis; PREVAILING SYSTEMS OF PHYLLOTAXIS. 39 ist and 2nd leaves, or between the 2nd and 3rd, is double that between the 8rd and 4th. These latter, it will be remembered, are separated by a long internode. The same order obtains with the succeeding whorls; the nodes, however, are now much more widely separated, while a true spiral arrangement, with the same and constant angular distance between all its leaves, is ultimately secured, and is henceforth continued uninterruptedly into the terminal bud, and represented by the fraction 2." Ап exactly analogous process takes place with opposite leaves in passing, for example, into the $ arrangement. Internodes аге, of course, developed ; and the leaves of the first few pairs that are separated from each other have not, if projected, an equal or correct angular distance of 144? between them successively ; but just as in the production of the $ spiral arrangement from whorls of threes, so is it here—that as the 2 plan is continued up into the terminal bud, the leaves “ right " themselves, as it were, and then ultimately come to obey this law accurately by securing a constant angular divergence. Fig. 140, on p. 168 of Sachs’s * Hand-book’ mentioned above, illustrates this fact, though that author has not detected the significance in the order of development of the leaves. The origin of Cycles.—It will be gathered from the foregoing remarks and from fig. 1 that « cycle is formed from a definite number of pairs of decussating leaves. "Thus, if opposite leaves give rise to the 2 arrangement, then the three pairs (0, 1) (2, 3) (4, 5) are required to produce the first cycle. Having once, as it were, “ started " it, the subsequent leaves are developed in accordance with the above law, the angular divergence between every two consecutive leaves being 144°. Similarly, if opposite leaves pass into the $ arrangement, then, in addition to the first three pairs, two others (6, 7) (8, 9) are required ; so that if the first cycle embracing leaves 0 to 7, inclusive, be secured, the subsequent leaves are developed according to the same law, requiring, however, 135° as their constant angular divergence. Hence it appears that while some plants have taken three pairs of the originally decussating leaves to form a cycle, i. e. of the $ arrangement, other plants have required five pairs, others seven, while cones frequently demand eleven pairs; so that the 21st leaf, or the first of the second cycle, falls over that of the first cycle, or that indicated in the figure as 0. Here, again, it does not seem possible to say why one plant, or even part of a branch, should thus convert a definite number of pairs of opposite and deeussate leaves into a cycle of some alternate arrangement. We can conceive of nature distributing the leaves of all plants in which they are alternate simply and solely іп accordance with the law given above, while it might have remained indifferent as to the precise amount of the angular divergence. ` As a matter of fact, however, this angle is determined by the and the fact that the point of emergence is determined, so to say, from within the axis before the leaf is formed seems fatal to the condensation theory. :- Again, Mr. Airy takes no account of stipules, which according to his theory ought to interfere materially in leaf- arrangement, because, if not petiolar (as in Rosacew) they rise from the axis exactly like leaves, and may even, as in the Elm and Lime, grow to as much as four times the size of the leaf to which they belong. Hence with, say, two minute leaf-apices and four subglobular or, what would be more true to nature, conical stipules, we may ask, What becomes of the “ contact tery " when the ionis development proves that the leaf-arrangement has not been interfered with? G2 40 THE REV. GEORGE HENSLOW ON THE ORIGIN OF THE number of leaves constituting а cycle, which cycle has itself been determined by the number of pairs of leaves of the originally decussate arrangement. It may be again observed that when numbers of the ** secondary " series, viz. 3, 1, 2, үү; &c., occur, as in the Jerusalem Artichoke, they do so from an analogous process arising out of whorls of threes. Further, in tracing the phyllotaxis of plants of paleeozoic ages, it appears that, besides representatives of the usual or primary and secondary series, some instances occur (as in species of Lepidodendron) which have resulted from the breaking- up of whorls of fours, giving rise to the tertiary series 1, $, 3, &c. ; while even a quaternary series, 1, 1, тг, «с., finds а representative in Knorria taxina, the phyllotaxis of which is said to be 25 *. At the present day, however, any other than the primary series is com- paratively rare. Correlation between Internodes and Angular Divergences.—If a stem grow rapidly, or at least develop internodes of considerable length (as is the case with the Jerusalem Arti- choke), the fraction representing the subsequent arrangement of the leaves upon it will be a “low ” one of the series. If, however, the internodes be very short, then, conversely, the leaves or their equivalents will be represented by one of the higher fractions, such as 125, S, 21, &c., as is well seen in cones, involucres, &c. There appears at present to be no better explanation of this correlation than the one already alluded to, namely the object of avoiding interference of one another. This, however, the author is inclined to regard as a distinct factor in the causes which regulate leaf-arrangement. It will be easily seen that if leaves are on the ? plan, with very short internodes, the 6th leaf will lie immediately upon the surface of the first, what would be called “ opposite” in flowers, especially if the leaves be sessile; but if they be on the т; or, better still, -%, then the 14th or 22nd leaf will be freer from the first, as seen in the scales of a cone. The Divergence зг determines all higher ones of the primary series.—In fig. 1 it will be seen that the following numbers occur in the same vertieal line,—0, 5, 8, 13, 16 [=2 х 8], and 21. Or, if written thus, 0, 5, 5 + 3, 8 + 5, 13 +3, 16 + 5, &е., the numbers сап easily be ascertained to any height required. If this be done, it will be | discovered that the numbers 34, 55, 89, 144, which form the denominators of the fractions higher than $$, do not fall in this same vertical line; and the question might be asked, on the supposition advanced in this paper, how it ever happens that higher fractions than 5) are represented in nature. The reply is, that while the numbers 5, 8, 13, and 21 give rise to the divergences 2, $, 1%, and 38; respectively, yet this last, зт, supplies all the higher fractions. That isto say, if the spiral arrangement represented by эт be projected as a helix, and the numbers noted down, to (say) the 90th leaf, then the 34th, 55th, and 89th leaf will be found lying nearest to the vertical line, alternately right and' left of it, while each of these leaves approximates nearer and nearer to it respectively, as will be seen by referring to fig. 2. Moreover, as the initial leaves of successive cycles of arrangement represented by the fractions y and higher ones, cannot be in a truly vertical line, it becomes often difficult to fix upon the leaf most nearly vertical; in fact it is almost, if not quite, arbitrary whether we select the 34th, 55th, or 89th leaf. We may go further, and say that the discovery or proof of a gene- ж Alluded to in Prof. Balfour's * Paleontological Botany.’ — OE PREVAILING SYSTEMS OF PHYLLOTAXIS. 4l rating spiral for any higher than $ is impossible; or perhaps it would be better to say that, for any other arrangement than those represented by the fractions 3, 3, 2, and 8, the phyllotaxis is more or less uncertain, and the idea of a definite or exact generating spiral becomes more or less evanescent. In the latter case the reason for hesitating to pronounee which is most nearly vertical over the first selected, arises from the fact that neither the axis nor the leaves or scales are ever developed with mathematical accuracy, notwithstanding the fact that the position of the leaves may be mathematically projected. It will, then, be seen how all apparently higher arrangements than +8; are due to that divergence itself; for this is the highest that can be determined from opposite leaves *. Moreover, to say that the angular divergence represented by зё furnishes all the higher ones, is, to some extent, if not inexact, yet a somewhat incomplete statement; for in certain cases ту has the same effect, since this fraction is also irrational to the circum- ference. Thus in some cones of the Scotch Fir, while the 13th scale is very nearly over ‚ the first, even the 34th leaf may not be far from the same vertical line, and so render it doubtful as to which fraction may best represent the phyllotaxis. However, it must be remembered that no higher denominator than 21 can be secured from opposite leaves, as stated above and shown by fig. 1. The reason why тү is the most usual kind of phyllotaxis with closely compacted scales seems due to the fact that i£ forms the nearest approximation to the alternation of whorled leaves, only one row of scales lying between those containing Nos. 0 and 21 respectively. | Theoretical considerations.-—Of course it is here only advanced as a theory that nearly all spiral arrangements of Dicotyledons, at least, have been originally derived from opposite leaves ; for although they have such in the cotyledons of their embryos, yet the very first leaves of the plumules (as is well seen in a germinating acorn) at once assume their own peculiar spiral arrangement. Yet at the same time it must be remembered that the theory is founded upon the fact that when a branch has opposite leaves below and alternate above, the latter are developed according to the law expressed in fig. 1. The law of embryology, that a member of a higher grade passes in its embryonic state through the conditions of a lower grade when in its adult state, would seem to hint at this being at least probable, and that opposite leaves were probably the original and normal state of the primogenitors of dicotyledonous angiosperms. . Moreover, when we remember that whorled leaves were apparently very frequent in the carboniferous epoch, and that instances of symmetrical reduction are common amongst plants, е. g. in the floral whorls of Circea as compared with Epilobium (in which * It may be said on à priori grounds that the failure of the theory in deducing any higher fractions than >; from opposite leaves is unfavourable to its acceptance; but it must be remembered that no theory is to be tested by what may be imagined it should account for, but by what it does. Now this theory, the author maintains, does account for all that is necessary ; and any assertion that 12 and higher fractions ought to arise directly out of opposite leaves is a merely subjective impression not necessarily requiring to be correlated with any objective fact. If they can be satisfactorily accounted for, аз the author maintains they ean, as stated, from the #- arrangement, then all that is required has been done. 42 THE REV. GEORGE HENSLOW ON THE ORIGIN OF THE latter genus, it may be observed in passing, there is a frequent change from opposite to alternate leaves), that the numbers * fours,’ * fives,’ and ‘sixes’ are often found amongst the whorls of the flowers on the same Jessamine, Elder, Rue, &c., and, again, that there is the positive fact of opposite and whorled leaves passing into spiral on the same stem, as in Epilobium, there would thus seem to be а just inference that the great majority of instances of spiral phyllotaxis have taken their origin from whorled and opposite arrangements. Opposite Leaves generally fail to produce the 4 and $ arrangements.—On referring to the figures of the diagram (fig. 1), it will be seen that for MIRA leaves to become spiral and distichous (that is, to be represented by the fraction $) the third leaf, No. 2, must, as it were, pass through 90? in order that it may fallover the first (or 0). Similarly if decussate leaves were to pass into the spiral arrangement represented by 3, 2. e. to become tristichous, the fourth leaf, No. 3, must fall over the first. In all other cases the leaves have to shift through a much less angle than 90°. This may be the reason why 4 and $ are by no means common amongst Dicotyledons. On the other hand, they appear. particularly frequent amongst Monocotyledons. Тһе above failure in producing the 1 and 1 arrangements is on the supposition that the spirals result from the conversion of а definite number of pairs of opposite leaves. Оп referring, however, to fig. 1, it will be seen that if leaf (0) be neglected, and No. 1 represent the first leaf, then the 4th is actually over the 1st, and thus will give rise to the $ plan. This may have been the origin of the tristichous arrangement amongst Dicotyledons, as well as of the ternary symmetry which occasionally occurs in the flowers of that class; but it is comparatively rare in both leaves and flowers; and this seems to correspond with the fact that it is usual to convert complete pairs into alternate leaves. Аз an additional faet it may be mentioned that other rare arrangements, such as 74, 74, $4, &c., which occasionally occur in nature, can be deduced from it, and lend proportional aid to the acceptance of the theory. This will be better seen by referring to the scheme appended to this paper (fig. 3). Before suggesting a eause for this latter fact, it would be as well to state that, in every case that I have examined, wherever leaves have to be represented by $ in Dicotyledons, I believe it is an effect produced by habit of growth resulting in an apparent production of the $ or $ arrangement. Thus, if the leaves on а shoot of Ivy growing against а wall be compared with those on a free (barren) branch, or the leaves of the common Yew with those of the * fastigiate" variety—or, again, comparing the leaves оп а horizontal bough of the common or Portugal Laurel with those on a shoot growing freely upwards, it will be seen at once how the usual and really or pice distichous arrangement reverts to the manner of growth represented by 2 or 2, thereby revealing the fact that in those (and probably all other) instances the leaves were not originally distichous, but have only acquired that arrangement by habit. l have, indeed, reason to suspect that it is only in those instances where the leaves of à branch lie in one and the same horizontal plane, from a peculiar habit of growth, that the $ and perhaps } arrangement occurs, and that this obtains mostly in а certain number of instances, where the leaves are broad and not dissected; for deeply gashed or com- PREVAILING SYSTEMS OF PHYLLOTAXIS. 43 pound leaves, as in the May and Ash, permit the ingress of light and air, so that a “ higher " spiral divergence is permissible, in that the leaves need not all be brought to the surface of the tree. As other examples of these divergences, may be mentioned the Alder, Lime, and Elm, the first of these being usually regarded as having its leaves arranged on the $ plan. This may also be due to peculiarity of growth, especially as the scales of both male and female catkins (of the Alder) are on the 2 plan. Again, if it be noticed how the Lime and Elm spread out tlieir leaves on horizontal planes, just as in the case of the garden-Laurel, the influence of light &c. already alluded to can be supposed to some extent able to have induced this arrangement ; how far, however, it is impossible to say ; for although light will cause fully developed leaves to turn to it, we do not know if it can cause a leaf to grow out at a particular point on the axis of the bud, so as to secure a more favourable position as regards the light, in anticipation, asit were, of its adult state. It is obvious that, of such broad leaves as these trees possess, if they were grouped spirally round the branch, some would fall underneath in the shade of the others, whereas by being, as they are, on the same horizontal plane, they all become equally favoured. Hence has arisen, as I believe, these distichous and, so to say, pseudo-distichous * and tristichous arrangements, which are by no means common amongst Dicotyledons 7. Having now seen that, whenever distichous and tristichous leaves are found amongst Dicotyledons, they may most probably be regarded as not issuing directly from opposite leaves, it may be observed, on the other hand, that these arrangements are particularly characteristic of Monocotyledons, as іп 18, orchids, grasses, &e. То ascertain a cause for this, let us again go to the embryo. Here we have but a single cotyledon. Now it must be observed that, for any arrangement of the usual or primary series, no portion of à spiral whieh when projected constitutes a circle ever contains more than three leaves ; or, with the exception of the distichous, such a circle always does contain three leaves. It will be at once seen that the distichous and tristichous, or the 4 and } plans, are the only kinds which admit of one complete cycle being projected in a circle. In the former the leaves are at opposite extremities of а diameter, in the latter at an angular distance of 120°. Hence there are these two methods of equalizing the distance between the leaves, and at the same time securing a cycle and satisfying the laws already stated. Thus, then, we may perhaps be able to account for the fact that $ and $ do frequently represent the leaf-arrangements amongst Monocotyledons. On the other hand, the con- Spicuous absence of arrangements represented by 2 and higher fractions is a significant : fact affording negative evidence of great value. * I make the distinction between distichous and pseudo-distichous, to indicate the fact that in some cases, as in the Yew and Tazxodium, the leaves are only apparently distichous in consequence of a twist in the petioles—a result more easily conceivable to be due to light than the truly distichous arrangement, as in the Laurel and Beech, in both of which the leaves or vertical shoots often revert to higher spiral arrangements. Т I take this opportunity of remarking that while examining the buds of these trees, I think I have discovered the cause of their unsymmetrical development—namely, that it is due to the position of the stipules, which lie obliquely across the two edges of the folded (conduplicate) leaf, thereby allowing one margin to develop more freely than the other. The bud-scales in both Elm and Lime are stipules. 44 THE REV. GEORGE HENSLOW'ON THE ORIGIN OF THE The relative Ages of the classes Dicotyledons and Monocotyledons.—1f any thing as to the relative ages of these classes can be argued from the foregoing facts, it would seem that Monocotyledons are of greater antiquity than at least angiospermous Dicotyledons, in that they furnish differentiated stages of leaf-arrangement issuing from an arrested condition of the embryo—that is, on the assumption of whorled leaves having preceded opposite, and opposite alternate, and that the single cotyledon of a monocotyledonous embryo has resulted from the arrest of another*. Тһе fragmentary character of many monocotyledonous groups, as witnessed by a greater proportion of isolated orders and genera t than in angiospermous Dicotyledons, as well as the very general prevalence of albumen (excepting water-plants), i. e. a retention of an embryonic character, may be brought forward as reminders that other facts corroborate this idea ; while, as far as the scanty evidence of geology can take us, Monocotyledons seem to have been represented in the primary epoch (though some paleontologists contest this), yet no instance of an angiospermous Dicotyledon has yet been forthcoming from the same period. Explanation of the Scheme (fig. 3).—In this scheme are represented the relative positions of the leaves of the ordinary phyllotaxis. The figures connected by double lines are those which give rise to Ute fractions 8; & ў, 1% С 21,4 2 Г» HE * In suggesting this process of reduction from whorled leaves as the origin of opposite in Dicotyledonous embryos, and of a single eotyledon in Monocotyledons, I may seem to differ somewhat from Mr. Herbert Spencer’s views expressed in his * Principles of Biology,’ vol. ii. p. 67, where he conceives of both Dicotyledons and Monocotyledons as diverging from a common ancestral type, the former having the first two folial or thalline organs flattened out in virtue of the requirement of strength by a development of the midrib, which therefore does not necessitate their being rolled into the form of a cylinder to acquire strength, as in Monocotyledons ; and, secondly, to use his own words, “if we assume, as the truths of embryology entitle us to do, an increasing tendency towards anticipation in the development of subsequent fronds—if we assume that here, as in other cases, structures which were originally produced in succession will, if the nutrition allows and no mechanical dependence hinders, come to be produced simultaneously, there is nothing to prevent Га passage from an , alternate to an opposite condition in the develop- ment of the first two thalline or folial expansions]. - If this ** assumption ” simply means * arrest" of the internode of the second frond, he is probably right, as it appears obvious that an alternation is aequired solely by the development of an internode. But the process of such internodal development appears rather'to be due to the presence of extra nourishment and rapid growth, than the “reverse, as seems to be implied by Mr. Spencer ; for his words intimate that nutrition must determine the opposition rather than the alternation. But I cannot help thinking that neither nutrition nor the want of it is the vera causa. They may determine more or less the growth but not the development, It is to me inconceivable how mere physical causes could have produced, for example, the fundamental distinction between an exorhizal dicotyledonous axial root and the endorhizal character of monocotyledonous adventitious roots—a feature which Mr. Spencer does not appear to _ have considered, and which is really due to the arrest of the primary or axial root in the latter, and which is there- fore correlated with an arrested cotyledon ; and, it may be added, an arrested condition is the presence of опен endosperm, so general in all terrestrial Monocotyledons. Hence I eonclude that an alternate or spiral arrangement in Dicotyledons results from the development of inter- nodes, but in Monoeotyledons in consequence of the arrest of one cotyledon. + I find from a calculation of the orders and genera given in Le Maout and Decaisne’s * Descriptive and Analytical Botany,’ edited by Dr. Hooker, that of Monocotyledons about 60 per cent. of the orders comprising that subkingdom have less than 6 genera, and nearly 33 per cent. have less than 3 genera; whereas of Monochlamydeous exogens, omitting Gymnosperms, there are about 40 per cent. of orders with less than 6 genera, and 26 per cent. with less than 3. Lastly, of Dichlamydeous exogens, there are no more than 28 per cent. of orders with less than 6 genera and about 15 per cent. with less than 3. : PREVAILING SYSTEMS ОЕ PHYLLOTAXIS. 45 No. 34 is placed vertically over 0. Strictly, however, this should be slightly to the left of the vertical line (but it would be inconvenient to allow for deviations due to minutes and seconds); for since 34 is not rational to the circumference, or a measure of 360, that leaf could never fall absolutely vertically over the first (0). The scheme may therefore be considered as approximately representing the 13 plan. The figures лоно by single zigzag 1 lines are those which give rise to the series 4 $, 6i to [= 81, 14 1%» tz == 8), 39, 33 $ [= 8 3$, «е. and which, as‘has been shown, are deducible from opposite leaves by omitting the lowest or first. There are several other fraetions which represent the angular divergences of leaves occa- sionally met with in nature, but which have hitherto not been referred to any phyllotaxis. Thus Hofmeister, in his * Handbuch der physiologischen Botanik, mentions the following :— forti p т. (15), Ай 209 $b 81 3517 s He endeavours to show how these may be derived font кодов of the ordinary series by a slight decrease or increase of the angular divergence, thus :— іі 35 or—i-— d т-і- Wy т=$+ ds rr =15+ 934 ОГ = $ — oy зб = $— зо ОГ = + d ЗЇ = 2 — тів ог = 34+ уз З= 3+ өз ог={— oy i? == 02—туув ог = тұ + біт. Ы. * To these may be added 1% and + Now, of мн ip P d 1i, oy ii сап һе more easily obtained from the new series, commencing 4 3 тт, &е., as given above. For 1% лї апа 15 p immediately from opposite leaves when the lowest leaf is neglected. Now 44 = 4 — zio з 7, and therefore differs from үр by a much smaller fraction than from $0ri. Again 11 differs from тг by only i+. So that while of the above ми fractions 2, 74 (which is simply the reverse of 1%), and 18 are deducible from some number or other respectively of the usual series, 3%, 11, and also 45 and H, are to be obtained by starting with fr, which, in turn, is directly obtainable from opposite leaves 1. * Excepting 14, these are indicated on the scheme by means of arrows. | t This fraction 1+ I cannot account for, unless it approximates the reverse of some other ; otherwise it cannot belong to the usual series, as there must be always 3 leaves in every projected circle of a cycle ; consequently 14 appears to be too high a number (of cycles) for 31 leaves. + Since this paper was in type I have met with a curious exception in Lagerstrümia, in which the order of the leaves, instead of forming a spiral, “ oscillate” through three quarters of a circle. The first four leaves succeeding the opposite are as in fig. 1; but the subsequent arrangement is as follows—O, 1, 2, 3, 5, ^ 455 8, 9, 19, IL, 13, 12, 15, 14, 16, 17, 18, 19, 21, 20. This, perhaps, may prove to be the clue to thè origin of distichous leaves amongst Dicotyledons. SECOND SERIES.—BOTANY, VOL. I. E © E E3 о ДА È Fa n © o 2 Е - 3 ЕЗ 22. 47 25 21 18 70 IV. On the Barringtoniacem. By Јонх Mars, F.R.S., F.L.S., Dignit. et Commend. | Ord. Imp. Bras. Rose, &c. (Plates X.-X VIII.) Read June 3rd, 1875. THE Barringtoniacee form an extremely natural group, offering very distinct and uniform characters; they differ from the Myrtacee in their alternate leaves without pellucid dots, in another kind of inflorescence, and in a fruit of very different structure. The inflorescence is either terminal and thyrsoid, with several approximated flowers, often of extraordinary dimensions, fixed upon a thick erect terminal stem, or more generally consisting of smaller flowers, spicately arranged upon long terminal pendulous racemes, sometimes subaxillary, or more rarely in branching panicles; but in all cases the flowers are formed upon one uniform plan. The leading character of this floral structure is, that their very numerous long slender stamens, in many series, are united at their base into a shortish erect monadelphous tube, . seated on the outer margin of a horizontal annular epigynous disk, from which they all fall off together, thus differing essentially from the Myrtacee, where the similar stamens are always free, each seated independently upon a perigynous annular disk, from which they fall off separately, leaving upon it many cicatrices, marking the several points of their previous attachment upon it. There is little analogy in this respect with the struc- ture of the Lecythidacee, where the short small stamens are separately supported upon numerous distinct processes, attached to a large petaloid organ (androphorum), of which no parallel is to be found in any other natural order. Another peculiarity of the Barringtoniacee is, that the tube of the calyx, agglutinated to an inferior ovary, is expanded above into a free limb, which in many cases offers the . very singular feature of being vesicoid, quite entire and undivided in the bud, marked by many parallel nerves, all meeting in a mucronate point in the apex ; after a while, by uniform internal pressure, caused by the growth of the petals and stamens, this limb becomes ruptured along the nervures, splitting into 2, 3, or 4 subequal segments, which eventually, little changed, are persistent, and crown the summit of the fruit. In other genera of the family the limb of the calyx consists of 4 free, small, rounded sepals, slightly imbrieated in sestivation; in a single instance it is cup-shaped, with an almost entire margin. The petals, constantly 4 in number, are generally 3 or 4 times as long as the calycine limb, are oblong, rounded, narrowing towards their base into fleshy claws, which are agglutinated to E. monadelphous tube of the stamens so firmly that all fall off together. H 2 45 MR. J. MIERS ON THE BARRINGTONIACEA. The stamens in several series are closely united at their base into а monadelphous tube; the filaments are slender, somewhat longer than the petals, and are spirally coiled in vestivation, but afterwards they spread into a feathery plume, are mostly furnished, at their curving apex, each with a small 2-celled anther; but in some genera those of the outer or of the inner series are bare, the intermediate series alone bearing anthers. The style, which rises from the centre of the ovary, is generally longer than the stamens, is subulate, rather slender, slightly curved, and terminated by a small elavate or lobed stigma, and is always persistent on the fruit. Another feature of importance is the disk, which assumes а form so peculiar that it — enables us to determine whether any plant presented to our notice belongs to Barring- toniacee or not—a test of much practical value. Аз before stated, it is epigynous, flat and annular, and bears upon its outer margin the staminiferous tube, while its inner margin is always more or less expanded into an erect urceolate prominence, sometimes into a tube of some length, which encircles the base of the style, leaving within it a deep hollow over the vertex of the ovary. "The ovary also affords good distinctive characters among some of the genera; it is always inferior, and adnate to the tube of the calyx; it is either 2- or 4-celled, with one or two, sometimes more, ovules in each cell. These ovules in some cases are erect, but more frequently are suspended from the summit of the cells by distinct funicles; or, when more numerous, they are attached to the axis; in all cases they are anatropous. In their subsequent development generally most of the cells are abortive, leaving only а, single fertile cell; and in that only a solitary ovule is matured. | The fruit varies in form, is generally of a large size, crowned by ће free persistent limb of the calyx; it has a thick. pericarp, frequently very thick, sometimes fleshy, but more often of a dry spongy consistence, lined within by an endocarp, forming an osseous or a coriaceous shell, covered by adhering longitudinal woody fibres. This shell is some- times 4-celled, but more frequently, by abortion, only 1-celled, with a single large seed; іп the genera Careya and Planchonia the seeds are smaller, more numerous, and im- | bedded in a tough pulp. In the African genus Petersia, the fruit assumes a very unusual form: the indehiscent pericarp is small, linearly oblong, thin, and scabrous out- side, and is cruciformly surrounded by 4 rounded membranaceous wings of much larger dimensions: this, by abortion, is 1-celled, containing a few linear terete seeds, attached to a parietal cord formed by its undeveloped three abortive cells. I did not succeed in determining the form or nature of the embryo, owing to the decay of the seeds. In Barringtonia and Planchonia the fruits are 4-celled, each cell in the latter producing several seeds, imbedded in a tough pulp. In these genera alone the seeds furnish an embryo with two distinct cotyledons, as will be subsequently shown. With the exception of these two and Pelersia, all the other genera (excluding one) uniformly present in each fruit a single large seed with an embryo of peculiar structure, upon the nature of which | ше opinions of botanists are much divided; and it is necessary to enter into much pre- iminary detail to ascertain the value of these conflicting views. This large embryo, apparently homogeneous in texture, invariably consists of two main | ке. опе реду encircled by the other at all points, often so intimately agglu- MR. J. MIERS ON THE BARRINGTONIACEJE. 49 tinated together that they appear like one homogeneously solid body ; but often, in dry- ing, they become visibly distinct, leaving a narrow vacant space between them. This structure was first demonstrated in 1791 by Gaertner’, who gave several illustrations of it. He regarded the outer body as albumen, and the inner one аз а pseudo-monocotyle- donous embryo, consisting of 2 fleshy cotyledons and a radicle, all intimately conferru- minated into a solid mass. : ; _ Blume, in 1826*, adopted this view ; and DeCandolle, in 18283, did the same. Roxburgh, іп 1832', accorded with the definition of Gaertner, in considering the nucleus to be formed of albumen, of 2 cotyledons and a radicle, all intimately agglutinated into . one solid mass; his description, however, is more important, as he figured and described the changes produced in the act of its germination. Within the pericarp he saw the inner body, which he regarded as the embryo, throw out a long shoot at each extremity, to form a new plant, the rootlet issuing from the summit of the fruit, while a young stem, covered with scales, forced its way through the bottom of the pericarp. He added that **this part, which was a prolongation of the plumular extremity, formed the ligneous centre of the shoots, while the external body furnished the cortical part and the leaves : in this growth there was no appearance of any cotyledon; he saw simply the 2 bodies regarded by him as albumen and embryo." He further explained this development by saying that * the albumen performs the same office as a cotyledon," repeating that ** by the elongation of its two extremities it furnishes the bark and foliage of the young plant, While the opposite end of the embryo, or central part, supplies the wood and pith, just as in Garcinia and Xanthochymus ; only there the central portion or embryo is very slender, and the permanent root proceeds from the base of the plumule, as in the Monocotyle- dones, while that from the opposite end of the embryo soon perishes, or remains slender compared with the other." Wight and Arnott, in 1834*, defined this embryo as formed of two concentrie homo- geneous combined layers, not separable into cotyledons an& radicle. Griffith, in 1835 °, described the seed of Careya, and its growth in germination. It consisted of an outer fleshy mass enveloping a central body ; theouter body is surmounted by a few fleshy scales concealing others; and these form part ef the plumule: аз the latter expands upwards, it carries with it the scales, which appear. alternating on the ascending stem, which, at its summit, throws out true convolute leaves; at first the scaly plumule is concealed by the lobules on the apex of the external body. It is evident, he says, that the inner subulate body is the root, and the minute scales the plumule ; so that the outer fleshy mass ought to be considered cotyledonary in its nature, “ and might be explained by supposing the cotyledons to be affixed in a peltate manner, and united into a solid mass.” This last passage is very obscure, and assumes what is improbable. "Wight in 18407 still inclined to his former opinion, and, in order to explain his views, copied from Roxburgh’s drawing the figure of a seed of Butonica as it appears when germinating within the pericarp. 1 уң &e. ii, 96, tab. 101. 2 Bijdragen, p. 1096. * Prodr. iii. 288. * Fl. Ind. ш. 634. Prodr. Fl. Pen. Ind. p. 333. € Proc. Linn. Soc. i. p. 280. ' Icon. tab. 152. 50 MR. J. MIERS ON THE BARRINGTONIACE. Lindley, in 1845 !, described the fruit of the Barringtoniacee as fleshy, and containing many bony seeds lodged in pulp, with an embryo in the axis of copious fleshy albumen. Wight, іп 18502, expressed his opinion that in Careya the embryo of its seed is placed in the axis of а copious fleshy albumen; and he figured it accordingly. Blume, іп 18522, repeated his conviction that in Butonica the embryo is exalbuminous, with the cotyledons and radicle intimately combined into one homogeneous fieshy mass, ignoring, as he had done before, that it is composed of two distinct concentric bodies, a fact adverse to his conclusion. Griffith, іп 1854, described the embryo of Butonica as formed of two conferruminated portions. He imagined that “in its central part the upper end represents a radicle, its lower end а plumule; he also supposed allthe parts which he figured, between the lowest scales of the superficies and the radicle, to be an adherent cotyledon, or else an immense radicle, and two or several minute cotyledons represented by scales, and an inconspicuous plumula—a peculiar form of embryo analogous to that of Dracontium and, in a less degree, that of Cryptocoryne.” The presence of these scales, however, was observed only in two species; probably in all the others they were absent. Griffith says distinctly they are absent in his B. conoidea, He evidently did not think they formed an essential part of the structure; for in the species he describes he noticed much irregu- larity in the number and situation of the scales. Miquel, in 1855, merely repeated Blume's definition in his monograph of f the family. Dr. Thomson, in 1857^, gave an able review of the opinions of botanists on this sub- ject, when he came to the conclusion that this form of embryo is exalbuminous, and con- sists of 2 concentric layers, that the cotyledons are rudimentary, that in germination the central layer is continuous with the pith, and the outer layer continuous with the bark of the new plant, that the plumule (at least almost without scales) is developed into the new stem, while the opposite extremity is elongated into a root. Не adds, that in this development the only appearance of foliary growth is in the series of minute scales upon the ascending axis, as shown in Roxburgh’s figure; but these are only rudimentary, for the first true leaves are not developed until that axis is 1 or 2 inches long. He had observed numerous instances of this germination, and adds, somewhat ambiguously, that the new stem is a prolongation of a bud springing from the axil of one of the minute scales observed by Griffith, He shows also the close affinity between this form of embryo and that of the Guttifere, as Roxburgh had before pointed out. Here I аш able to offer some new evidence bearing upon this subject. Іп 1854’, in my paper “ Ор the Structure of the Seed and peculiar Form of the Embryo іп Clusiace@,” I showed that it consisted of 2 layers, one placed concentrically round the other; I gaye to the outer one the name of exorhiza, and to the inner one that of neorhiza, because, in germinating, the latter emitted at one extremity a growing stem, at the other a new rootlet; and in confirmation of this view I copied from Roxburgh’s drawing a figure of * Veg. King. p. 754. 2 Ilustr. p. 20, tab. 100. 3 In Van Houtte, Flor. Serr. iv. p. 72: | * Notule ad Plant. Asiat. iv. pp. 657, 658, pl. 636. * Flor. Ned. Ind. i. 484. * Proc. Linn. Soc. ii. 52, tab. 1. * Linn. Trans. xxi. 243, tab. 26. MR. J. MIERS ON THE BARRINGTONIACE.E. 51 the seed of Xanthochymus in germination’. I had not then witnessed the fact myself ; but not long afterwards Dr. Spruce kindly sent me the seeds of a Clusia which he gathered in the act of germination and in its several stages of growth. Неге the central body I had described became much broader below, and had forced its way through the apical lobes I had figured, rising in a naked subulate form, till it had grown to 10 times the length of the outer body, which had not enlarged in any way, but which, on the contrary, had withered into a mere membrane, after yielding its substance to nourish the growing portion. Хо scales were visible on the rising shoot; but when it had reached the height above mentioned, it threw out at its summit at first two, and then other decussating pairs of broad rounded leaflets; and at the bottom of the seed a distinct rootlet was evolved. This agrees perfectly with the figure of Roxburgh; only there were no scales, and no appearance of а second root close to the plumule. Тһе specimens of the Clusia-seeds are. now before me; so that 1 can vouch for the correctness of this description. From this we may gather the important fact that the outer body of the embryo forms here no portion of the new plant, and that its use appears only to perform the part of а cotyledon, in affording nourishment to the inner growing portion. It is probable that the same occurs in the germination of the large fleshy embryo in Butonica, which Rox- burgh figured as growing in the same manner?; but we have no evidence to show what becomes of the outer exorhizal portion in those cases, or that it forms any part of the new plant. This would lead to the conclusion that in Butonica the exorhiza is simply a cotyledon under a sheathing form, analogous to the radicle of an ordinary dicotyledonous seed from which the cotyledons have have been eut away, leaving a cylinder consisting of an outer sheath confluent round an inner body, terminated by the plumule. An example somewhat similar is figured by Gaertner in Hippocastanwin *, with this distinction, that the plumule is there coleorhized, a very rare circumstance in seeds eminently dicotyledonous. One important point of structure was noticed by Dr. Thomson in the речин of Buto- nica: in а section of it, under the microscope, he found that both the exorhiza and neorhiza are alike formed of simple cellular tissue and starch granules, but that between them, adhering to both, there exists a distinct layer of delicate vessels, continuous up- wards and downwards with the ligneo-vascular cylinder of the stem and root*. This, without doubt, is the medullary sheath described by Mirbel®, investing the cellular tissue of the axis of the plumular support, and forming there all the vessels which con- stitute the wood and bark in the new plant. : . А very instructive example of the kind of embryo іп Butonica occurs in that of Caryocar, which has a gigantic amygdaloid neorhiza, surrounded by a rather thin exorhiza of uniform thickness; the former is of a crescent form, in the upper horn of which is observed the radicular point, and in the lower horn the plumular extremity. The germination commences within the pore» ‘аз soon as the seed is formed, by the ' Linn. Trans. xxi. tab. 26, fig. 34. 2 Wight, Icon clii. fig. 5, and Roxb. Flor. Ind. ii. 635. * De Fruct. ii, 135, tab. 111. fig. б, Н. 4 Journ. Proc. Linn. Soe. ii. pp. 51, 52. 5 Elém. i. 114. 52 МВ. J. MIERS ON THE ВАВВГУСТОМ1АСЕХ. plumule perforating the thin coating of the exorhiza, and protruding from it in the form of a collet, bearing 2 very small lobes; this is done while it is confined within its entire integument; but the growth is of short duration, being arrested by the strong resistance of the bony nut, while the inward pressure causes the exorhiza to become moulded into the channel in which the plumule lies. Тһе lobes just mentioned were regarded by Gaertner as cotyledons; and so they have been universally considered; but any one who will reflect for a moment must see that they are, in fact, the plumule; for Gaertner says he could discern no other, and his expression “plumula nulla" is an obvious incon- sistency. We may obtain, by means of analogy, some additional light by comparing the embryo of Butonica with that of Rhizophora, which germinates also within the pericarp. There the ovary is 3-celled, with several ovules; but the fruit, by abortion, is 1-celled and monospermous; the pericarp is oblong, about 1 inch long, coriaceous, partly inferior, mostly superior, terminated by a bifid style, and surrounded above its base by the per- sistent calyx ; the seed which fills its cavity is invested. by a thin integument, has no albumen, and an embryo of singular development, consisting of an external exorhiza and an internal neorhiza. The latter, while upon the tree, begins to swell, and forces its way through the integument and through the pericarp, by rupturing a small hole between the lobes of the style, and it grows to a length of 10 inches by the time that the exorhiza has only extended half an inch beyond the pericarp. This exorhiza is 1 line thick, is of a greenish hue, and is filled with soft threads, probably oil-cells, such as I described in the embryo of Clusia, its upper extremity being lacerated by the protrusion of the neorhiza, while at its base, within the calyx, it is terminated by a hollow dome of soft yellow con- sistence, containing no threads; except in this part, it is confluent with the basal portion of the neorhiza, but is free from it at the dome, which encloses the free plumule that terminates the end of the neorhiza and consists of 4 or 6 small leaflets plicated, con- voluted, and converging to a point. This dome-shaped portion is called the calyptra by Jacquin, the albwmen by Gaertner! ; the exorhiza is the erus of J acquin, the vitel/us of Gaertner; the neorhiza is the semen of J acquin, the radicula of Gaertner; while the ter- minal free end of the neorhiza is the conus of Jacquin, and the cotyledons of Gaertner. Jacquin gives a long account of its germination and subsequent growth. The semen (neorhiza) becomes the trunk of the new plant; the tubercles on its surface throw out numerous rootlets that form, ultimately, arched buttresses for its support on the sea-bed beneath it; while, at the other extremity, as soon as it can detach itself from the peri- carp, the plumule expands into an ascending stem, crowned by permanent real leaves. In this development. the exorhiza seems to perform no other part than to afford its nou- rishing juices to the young plant; and it gradually withers, as in Clusia; the medullary sheath coating the neorhiza quickly generates a system of woody vessels, so hard as to render it difficult. to be cut by a knife, the neorhiza being contracted into a narrow | central pith,—a structure confirmed by the observations of Griffith ; the embryo may be said to be coleorhized by the concealment of the plumule within the terminal dome, * Griffith (Notule, iv. 664) calls this cotyledon. | MR. J. MIERS ON THE BARRINGTONIACE X. 53 which bears much analogy to the nipple-shaped process described by me in Clusia, арра- rently divisible into 2 or 4 lobes, which formerly I regarded as rudimentary cotyledons’. The kind of solid embryo found in Butonica and many other genera, under different modifications, consists, as before said, of an external exorhiza and an internal neorhiza, the latter having a plumule at one extremity and a radieular nipple at the other, the ends expanding in opposite directions in germination, the one being the precursor of the ascending stem, the other of the descending root ; between the two main fleshy bodies, and adhering to both, is the medullary sheath’, consisting of elementary vascular tissue, which, by the nutriment afforded on either side, yields woody fibres to the root, and gives origin to the wood, bark, and leaves of the new plant, the neorhiza, void of vessels, remaining as the central pith, while the exorhiza, also without vessels, merely gives out all its substance as nutriment to the general growth, and, without expansion, gradually dies away, thus performing simply the function of a cotyledon. То avoid frequent repe- titions in the following descriptions, this kind of embryo will be said to be mesopodal, because the expansion of its growth in germination is always in the axial portion (neo- rhiza), while the outer portion (exorhiza) finally becomes atrophied. The name tigellwm has been given to the analogous kind of embryo in Xanthochymus* and other Guttifere ; but it appears to me a very objectionable term, because the word tigella, invented by Mirbel in 1815*, and since used by all French botanists, denotes that portion of the plumule, in ordinary cases, which lies between it and the neck of the radicle: this, however, is seldom visible till after the commencement of germination. Jussieu also used it in the same sense?; but in a previous page he applies it to the great fleshy mass in the embryo of Pekea^. Consequently the name /ige//wn ought not to be given to an entire embryo of peculiar form when the same designation has been extensively used by botanists to denote a very small portion, and that often invisible, in the ordinary forms of the embryo in the seeds of both exogenous and endogenous plants. Such are the leading characters of the Barringtoniacee, from which it will be seen that they differ widely in their structure from the Муг/асев, and equally so from the Lecythidacee. With the former they have hardly a single character in common ; for the apparent similarity of their numerous long stamens is destroyed by the insertion of each filament separately upon а perigynous disk, while in the Barringtoniacee they stand in many dense series, all united at their base into a monadelphous tube, inserted on the outer margin of a distinctly epigynous disk. From the Lecythidacee they differ in a totally distinct floral structure. The Barringtoniacee appear to me to approach nearer to the Rhizophoracee; indeed, if we imagine in any plant of the former a flower with the staminiferous tube aggluti- tinated to the limb of the calyx, it would at once be referred to the latter family ; it would be especially close to the new genus На’тепа, established by me upon а plant ! Linn. Trans. xxi. 246, tab. 26. figs. 8-10, 26, 27. * Mirbel, Elém. i. 110; Juss. Cours Elém. p. 48, fig. 88, 99. 3. Ann. Sc. Nat. sér. 4, vol. xiv. tab. 17. fig. 15; Trimen, Journ. Bot. 2nd ser. vol. iv. р. 67, tab. 160. * Elém. pp. 50, 601, pl. 57. * Cours Elém. p. 365. fig. 446, p. 366. fig. 447. * Loc. cit. p. 350. fig. 420 (in adnot.). SECOND SERIES.—BOTANY, VOL. I. I 54 MR. J. MIERS ON THE BARRINGTONIACE. from Fiji, collected by Dr. Seemann, and which, singularly enough, he referred to Bar- ringtonia speciosa’. In thus alluding to Rhizophoracee, I mean the order as established by the late Mr. Robert Brown’, excluding the Legnotidee and Cassipouree*, since united with it‘, but which, it seems to me, ought to be located elsewhere, because they differ widely in the habit of the plants; and in their yielding a resinous juice, in the presence of stipules, in a more panicular inflorescence, in their flowers with laciniated or fimbriated petals, in their semi- or wholly superior ovary, in their fruit containing numerous seeds imbedded in pulp, each seed with a terete dicotyledonous embryo enclosed in copious albumen. There seems, indeed, but little intimate relationship be- tween them, while the true Rhizophoracee form a well-marked family, the uniformity of which is destroyed by its association with the groups before mentioned. The Barringtoniacee have no representative in the American continent, their exist- ence being confined to the Old World, over the tropical portions of which they аге widely spread, extending from Africa throughout the Indian and Malay peninsulas, growing in the numerous islands of the Malayan archipelago, in Australasia, in the broadly dispersed islands of the Pacific Ocean, but never reaching the continent of the New World, where the Lecythidacee exclusively occupy their place. This is. a fact of some value in the question of the geographical distribution of plants, and powerfully tends to support the opinion of those who contend for the separate origin of distinet types. The genera already described by botanists are Barringtonia, Butonica (which I have restored), Stravadium, Careya, Planchonia, and Petersia, to which are now added Agasta, Пототта, Megadendron, and Chydenanthus, all marked by peculiar characters, as the following clavis will show :— GENERUM DISTRIBUTIO. A. Calycis limbus in alabastro indivisus, maximus, oblongus, demum in lobos 2 ruptus. a. Embryo dicotyledoneus: semina plurima . . . . . . . . . . . . 1. BARRINGTONIA- б. Embryo mesopodus: semen solitarium . . . қ ‘ . 2. ÁÀGASTA. B. Calycis limbus in alabastro indivisus, mediocris vel а onm in aos 8—4 ruptus. Embryo mesopodus: semen solitarium . . . . . . 3. BUTONICA. C. Calycis limbus in alabastro sectus. Sepala 4 imbricata. c. Ovarium 2-loculare. * Embryo mesopodus: semen solitarium magnum, pericarpium siccum. . 4. STRAVADIUM. d. Ovarium 4-loculare. ** Embryo dicotyledoneus: semina plurima in pulpa nidulantia . . . . 5. PLANCHONIA. *** Embryo mesopodus: semina plurima in pulpa nidulantia . . . . . 6. CAREYA. **** Embryo mesopodus : semen solitarium, magnum; periearpium siccum . 7. DoxomMa. *x*** Embryo ignotus: semina plurima, linearia; pericarpium siccum, in alas 4 maximas membranaceas expansum . . . . . . . . =. . . . 8. PETERsIA. D. Calycis limbus sectus: sepala 3—4, valvata. Embryo mesopodus: semen soli- tarium, magnum; pericarpium siccum . . . . . . 9. MEGADENDRON. E. Calycis limbus integer, cupularis: inflorescentia опа а . . . 10. CHYDENANTHUS. * Flor. Viti. i. p. 82. * Gen. Remarks, p. 549. * For reasons given by Blume (Mus. Bot. Lugd. i. p. 126). * Endl. Gen. p. 1186; Benth. Hook. Gen. Pl.i.p.678. ` MR. J. MIERS ON THE BARRINGTONIACEA. 55 1. BARRINGTONIA. This genus was established by the two Forsters in 1776, after their return from the second voyage of Capt. Cook, when they published very good analyses of its floral and carpological structure ', the plant itself being well represented in the drawings of Forster and of others who copied the same. The younger Linnzeus, іп 1781, gave its generic character * and details of the typical plant?, obtained from a communication made to him from Forster ; but as the plant and drawings of it were both unknown to him, he wrongly con- cluded that it was identical with Osbeck’s Mammea asiatica, with Sonnerat's Commersona, and Rumph’s Butonica. Blume іп 1827%, perhaps wholly unacquainted with the details and drawings of Forster, applied the name of Barringtonia speciosa, as the younger Linnzus had done, to other plants found by him in Java; hence in his generic character of Barringtonia he ascribed to it carpological features quite at variance with Forster’s description and analysis; and to this cause we may attribute all the complications that have since occurred. DeCandolle, in 1827 °, not perceiving this mistake, adopted Blume’s definition uncon- ditionally—an example followed by Roxburgh, Wight, the Dutch botanists, and authers of every nation since that period; and hence the general confusion at present existing, from which I have here attempted to extricate the family. One of the most remarkable characters of Barringtonia is the presence of a dicotyle- donous embryo in the seed, a unique occurrence till lately, when the genus Planchonia was established by Blume himself? upon some plants from the same region, which present an embryo similar to that of Barringtonia. There can now, therefore, be no reason for refusing the acceptance of Forster’s characters. As yet we know only the typical species; and it is remarkable that no one except Montrouzier appears to have seen the plant since Forster collected it; the only specimen of it extant, as far as Гат aware, is the original type, fortunately preserved in the British Museum. Sir Joseph Banks purchased some duplicates of Forster's plants; but no specimen of the fruit appears in the Banksian collection. It is therefore almost wholly from Forster’s materials that the following generic character has been framed, a few further particulars being furnished by Montrouzier. BanniNGTONIA, Forst. (non alior.). Calyx in parte adnatus, limbo magno, in alabastro oblongo, integre clauso, parallele nervoso, demum in lobos 2 concavos persistentes rupto. Petala 4, magna, oblonga, unguibus ad tubum staminigerum agglutinatis. Stamina numerosissima, pluriseriata, in tubum cylindrieum subbrevem monadelpha, et cum petala conjunctim caduca : filamenta subfiliformia, in estivatione spiraliter ботой; demum recta, dilatatim expansa, petalis longiora, colorata; anthere subparve, dorso basin versus affixe, ovato-rotundate, flav:e, 2-lobe, lobis adnatis longitudinaliter dehiscentibus. Discus epigynus, horizontaliter annularis, pulvinatus, margine externo tubum staminigerum fulciens, margine interno in eminentiam longiusculam anguste tubularem apice dentatam stylum cingentem expanso. Stylus subulato-teres, filamenta paullo excedens, persistens. Stigma parvum, capitato-oblongum, cavum. ` Char. Gen. p. 75, tab. 38. Ё. a, b, c. ? Linn. fil. Suppl. p. 50. * Ibidem, p. 312. * Bijdr. p. 1096. * Prodr, iii. 288. * In Van Houtte, Fl. Serr. vii. p. 24. I2 56 MR. J. MIERS ON THE BARRINGTONIACE М. Ovarium inferum, turbinatum, vertice circa stylum angustissime concavo, 4-loculare ; ovula in quo- que loculo 3, imo radiatim affixa. Fructus magnus, obverse pyriformis, apice sensim attenuatus et calyce coronatus, obsolete 4-gonus, indehiscens: pericarpium crassissimum, extus Ігеуе ; mesocarpium crassissimum, subcarnosum; endocarpium durum, subosseum, extus fibris rugose lignosis laxe intricatis tectum, intus leve, 4-loculare: semina in quoque loculo solitaria, subglobosa, rugosa; embryo conformis, exalbuminosus ; radicula ejus longitudine, teres, incurva, apice sursum spectans, truncata, imo brevissime incurvata, et hine cotyledones 2 magnas, erectas, crassifoliaceas, parallele plicatas, æquilongas suffulciens. Arbor in insulis Maris Pacifici vigens ; folia majuscula, cuneato-oblonga, brevissime petiolata, glaberrima ; inflorescentia terminalis, racemosa ; flores plurimi, speciosi. BARRINGTONIA SPECIOSA, J. В. & G. Forster, Char. Gen. p. 76, tab. 38. f. а, 6, & с (1776); G. Forst. Icon. ined. vol. ii. tab. 191 (1776); J. F. Miller, Icon. ined. fasc. ii. tab. 7 (1776); Linn. fil. Suppl. р. 312 (1781); Cook's Voy. vol. i. tab. 24 (1784); С. Forst. Prodr. Fl. Austr. р. 47 (1786); Kerner. Hort. Semp. vol. i. tab. 28 (1796) ; Guillem. Zeph. Tait. Ann. Sc. Nat. 2° sér. vol. vii. p. 358 (1837), ex MSS. J. 6. Forster ; АИ іп Mém. Acad. Lyon, vol. x. р. 309 (1858): Butonica, Lam. (in parte) Diet. i. 521; Illustr.tab. 590 (non 591), ex icon. Mill. iterat. (1783): Butonica speciosa, Dryand. in Aiton, Hort. Kew. (in parte) ii. p. 439 (1789): arbor procera, trunco crasso, erecto, ramosissimo, cortice cinereo, fusco, rimuloso ; ramulis patentibus, crassiusculis, fistulosis, rugosis, pallide brunneis, sulcato-striatis, apice confertim foliosis, inferne (e foliis lapsis) late cicatricatis: foliis obovatis, apice sensim obtusatis et smpissime emarginatis (junioribus acutioribus), infra medium subcuneatis et in petiolo lato brevissimo decurrentibus, hine fere sessi- libus, patentibus, quam latitudo duplo longioribus, integerrimis, subcoriaceis, supra profunde viridibus, siccis fuscis, nervis remotis, flavidis, paullo divaricatis et arcu- atim nexis, costam versus prominulis, subtus paullo pallidioribus, costa promi- nente, imo sensim incrassata, et rubicunda: inflorescentia terminali, erecta, thyr- soidea, 3-20-flora; floribus amplis, speciosissimis, alternatim evolutis; rachi crassa, subangulata; pedicellis subsparsis, subpatentibus, longis, validis, apice incrassatis, imo bractea foliosa cuneato-rotunda integra sessili munitis; calycis limbo majus- culo, ovato-globoso, apice mucronsto, integre clauso, pedicello ter breviore, demum in lobos 2 :equales parallele nervosos coneavos patentes rupto; petalis 4, triplo lon- gioribus, obtuse oblongis, expansis, candidissimis, marginibus revolutis, imo roseo- tinctis, unguibus tubo staminifero agglutinatis; staminibus numerosissimis, imo in tubum monadelphum coalitis; filamentis petalis duplo longioribus, divergentibus, candidissimis, apice roseis; antheris parvis, luteis; disco epigyno, late pulviniformi, margine externo tubum staminigerum fulciente, interno іп tubum angustum егес- tum longiusculum apice denticulatum styli basin vaginantem expanso; stylo subu- lato, tenui, candido, apice sanguineo, staminibus vix longiore; ovario infero, turbi- nato, 4-loculari, ovulis in quoque loeulo 2-3 ab axi radiantibus: drupa maxima, | obpyriformi, calyce coronata, superne сопісе angustata, inferne rotundata, sub- 4 4-gona ; регісагріо levi, viridi, ѕіссо rufo-fusco, crassissimo, solide carnoso; endo- E carpio osseo, mueiformi, ovato, extus fibris validis lignosis tecto, intus 4-loculari, loculis ssepius monospermis ; seminibus ovato-globosis, rugosis, cerasi magnitudine, МБ. J. MIERS ON THE BARRINGTONIACE.E. 51 embryone dicotyledoneo generis. In insulis Societatis, Amicorum, et Nova Cale- donia : v. pl. s. in herb. Mus. Brit. Tahitia, Tonga (Forster, pl. typica); fructus non vidi. The above imperfect specimen from the Banksian collection is fortunately preserved in the British Museum. The leaves in this specimen and in Forster's drawings are 7-103 in. long, 99-5 in. broad, and almost sessile, are crowded at the end of the branches, scarcely 3 lines apart. The terminal raceme is from 6 inches to near a foot long; its thiek erect rachis bears from 3 to 20 flowers successively developed, the upper ones (in bud) being only one third of the size of the lower or full-grown ones; the pedicel when full-grown is 5 in. long, 14 line thick below, increasing to 3 lines in thickness, at the base of which is a foliaceous bract 2 іп. long, 1 in broad. The limb of the calyx in the full-grown bud, is quite closed at the summit, is 4 in. long, 9 lines in diameter, and splits open into two large concave expanding lobes, which are marked by strong parallel nerves, and, with the style, are per- sistent on the fruit; the petals are pure white, 14 in. long, 1 in. broad, and reflected ; the Stamens are 3 in. long, united at their base into a tube 4 lines long, 6 lines in diameter ; the style is the length of the stamens; the turbinated 4-celled ovary is 6 lines long; the epigynous disk, crenulately pulvinate, is 6 lines in diameter, bearing on its outer margin the staminiferous tube, while its inner margin is prolonged into an erect tube 2 in. high, narrowing upwards into a pluridentate open mouth, thus concealing the lower part of the style, a section of which is shown in Miller’s drawing. The fruit, exclusive of the calycine limb, is 4 in. long and 34 broad below the middle. Тһе ovate nut is 4-celled, each cell eontaining a single seed, which in Forster's drawing is 8 lines long, 6 lines broad, showing a terete radicle of its whole length, turned up at base to support two - erect fleshy plicated cotyledons, as I have seen in Planchonia. This plant (which is here shown in Plate X., taken from Forster's drawings) must not be confounded with Solander’s species, nor with that of Linnzus, as it has hitherto been. The fruit is called Futu by the natives, and, like many others of the same family, is poisonous and used to stupify fish in order to catch them. Seemann mentions (Pl. Vit. p. 83) the fruit of a similar kind, and probably another species, called in Fiji Vutu-dina (genuine Vutu). He did not see it; but it was said to be larger, its exterior portion soft, containing а nut so hard as to require a sharp instru- ment to open it in order to get to the seeds, which are edible, thus differing from Forster's species. This may probably be the species described by Montrouzier under the name of Barringtonia speciosa, which he found in the island Art, near New Caledonia. Although he mentions nothing of the shape of the leaves nor the ; character of its inflorescence, he gives a full account of its floral structure, which is similar in general details to Forster’s plant, but differsin its petals, which are not longer than the calycine lobes; the stamens, too, are shorter; the petals, though white, have a roseate hue, with their margins revolute; the inner margin of the disk seems to differ in being shorter and more urceolate in form, and fimbriate (not dentate) on its margin; the ovary is t-locular; and he noticed at the base of three of the cells 4 ovules. To the plate of Barringtonia speciosa in Miller's showy book no information is attached, 58 MR. J. MIERS ON THE BARRINGTONIACE. and it is little more than a copy of Forster's drawing; so also is the plate given in the narrative of Cook's first voyage, and in a wrong place, with an apology from the editor for the omission of any details. Тһеге are some circumstanees in the history of this species in relation to the two Forsters that are deserving of notice. John R. Forster, a botanist of some standing, and a friend of Linnzus, was appointed to the * Resolution' in 1772, on the second voyage of Cook, as botanist to the expedition, with the assistance of his son George, then only seventeen years of age, as draughtsman. After three years’ absence they returned to England in 1775. Ц was previously agreed that the two Forsters should describe the history of the voyage, the profits of which were promised to them. On presenting the first portion, the Lords of the Admiralty would not sanction it unless many obnoxious passages were expunged. This the Forsters refused to do, and withdrew in disgust, the elder to Germany in 1776, where he was appointed Professor of Natural History at Halle, remaining there till he died іп 1798. His son George at the same time retired to become Professor at Cassel, and afterwards to Mayence as librarian to the Elector; here entering into the excitement of the French revolution, he went to Paris as deputy to the National Assembly. During his absence the Germans stormed and took Mayence, when all his books and papers were lost, and, bereft of all his property, he left France in despair and went to India, where he died in 1794. Bearing in mind these dates, we find that Linnzeus published the diagnosis obtained from the younger Forster a year after his return from his voyage, and in the same year that Forster’s ‘Char. Gen.’ was published. In that work the new genus Barringtonia has as its synonym the Butonica of Rumph, tab. 114; that was adopted by Linneus with the addition of another synonym, Mammea asiatica, Linn. It is worthy of remark that in that copious diagnosis no notice is taken of the singular nectary, which forms one of the most peculiar features of the genus ; on the other hand, in the details of the species, said by Guillemin to be in the handwriting of Forster, this nectary is fully described, and the synonyms above mentioned are also given. Тһе character of the fruit, especially of the seed, as given by the Forsters in their ‘Char. Gen., is there wholly suppressed, and in its place a description of that of Rumph’s Butonica is substituted. Another reference is given, to DC. Prodr., published many years after the death of the two Forsters. We may therefore infer that there has been a little tampering with the original MS. in both cases, apparently to suit the views of other persons. Kerner in 1796 published a beautiful drawing of Barringtonia speciosa, copied from that of G. Forster; but he added to it the fruit of Rumph’s Butonica, tab. 114, instead of the fruit and seed so clearly figured by the two Forsters. This plate is accompanied by а description copied from that of the younger Linnzeus. Van Houtte (Fl. Serr. vol. iv. tab. 409) published coloured drawings purporting to represent Barringtonia speciosa; but these were copied from Paxton’s without acknow- ledgment ; this I have noticed more particularly under Agasta asiatica. Four. years later he published (op. cit. vol. vii. р. 23) a monograph of the Barringtoniacee, as- sumedly from the pen of Blume, where, beginning with B. speciosa, he gives to it a whole page of synonyms and references belonging to numerous other species, including НИ РАЕН ИРИНА ETT MR. J. MIERS ON THE BARRINGTONIACE.E. 59 many of Butonica and Stravadium confounded with it, a list which I feel assured never received the sanction of Blume. The before-mentioned form only a portion of the many complications in which Barringtonia speciosa has been involved by authors; several others are exposed in the sequel. 2. AGASTA. A few of the species referred by authors to Barringtonia differ from that genus so essentially, and are so uniform in their structure, that I propose to unite them into a new genus, Agasta'. I have selected as the type the magnificent species first collected by Dr. Solander in 1769 at Otaheite, when he accompanied Capt. Cook in his first voyage round the world. This plant, scarcely known to botanists, is well represented in Parkinson's inedited beautiful drawings, and is copiously described by Solander in his unpublished notes preserved in the British Museum, where he gives a full aecount of its floral and carpological characters. The genus agrees with Barringtonia in its large thyrse-like raceme of splendid flowers, in its extremely large vesicoid calyx, entire and undivided in the bud, subsequently splitting into two lobes by the pressure of the growing petals and stamens; but it differs from that genus in the development of its large fruit, and the entirely dissimilar structure of its seed. Solander, in his notes, gives an exact detail of this development, showing that the seed is much larger, occupying the entire space of its single fertile cell, being egg-shaped, consisting of an exalbuminous solid embryo, of mesopodal structure, as before described (ante, p. 53), and which is found in most of the genera of this family. AGASTA, nob. Barringtonia auct.; Butonica, Soland. (non Rumph.) ; Mammea (in parte), Linn. Calycis adnati limbus maximus, ovato-oblongus, parallele nervosus, in alabastro integer et omnino clausus, demum іп lobos 2 concavos ruptus, in fructu persistens. Petala 4, lobis calycinis 3—4plo longiora, euneato-oblonga, unguibus tubo staminigero affixa et cum illo caduca. Stamina numero- sissima, pluriseriata, imo in tubum brevem monadelpha ; filamenta filiformia, petalis longiora ; anthere parve, bilobe, lobis adnatis, paullo supra basin dorso affixis, rima longitudinali dehiscentibus. Stylus filiformis, longitudine staminum, persistens; stigma parvum, simplex. Discus epigynus, pulvinatim annularis, margine exteriore tubum staminiferum fulciens, interiore in urceolum brevem erectum productus. Ovarium inferum, turbinatum, vertice intra discum circa stylum cavato, 4-loculare; ovula in quoque loculo 2 vel plura, apice funiculo brevi suspensa. Fructus majusculus, conice vel depressius obpyriformis, 4-gonus, levis, calyce coronatus: pericarpium crassissimum, spongioso-carnosum ; endocarpium coriaceum, extus fibris numerosis lignosis tectum, intus abortu l-loculare, cum loculis marcidis 3 in chorda longitudinali prominente eo affixa signatum, mono- spermum : semen magnum, loculo conforme; testa submembranacea, endocarpio adhzsa; embryo ' exalbuminosus, solidus, eburneus, chords impressione longitudinaliter canaliculatus, esopodus. Arbores Asiatice, aut in insulis Oceani Pacifici vigentes, alie, ramosissime : folia maxima, т apicibus ramorum conferta, oblonga, breviter petiolata, alterna: thyrsus terminalis: flores speciosissimi, alterni: pedicelli longi, apice incrassati, imo bractea foliosa decidua muniti : fructus viridis, siccus brunnescens, nitidus. | ` From áyazrós, admirandus. 60 MR. J. MIERS ON THE BARRINGTONIACEJE. 1. AGASTA SPLENDIDA, nob.: Butonica splendida, Soland. Prim. (ined. 1769), p. 282!; in Obs. MSS. Soland. (ined.), p. 382 (excl. synon.); Parkinson, Icon. ined. vol. ii. tab. 68 et 69: Butonica speciosa, Dryand. in Aiton, Hort. Kew. 15% edit. (1789), vol. ii. 439 (excl. syn.): arbor elegantissima, ramis late expansis: foliis maximis, sparsis, patentibus, latissime obovatis, supra medium hemisphericis, basin versus cuneatis, in margine cartilagineo integerrimis, subcoriaceis, planis, in junioribus ovalioribus obtusis et sspe emarginatis, supra nitidis, intense viridibus, siccis pallidioribus, nervis paucis valde remotis prominentibus, subtus pallidioribus, nervis venisque retieulatis paullo prominentibus, costa imo incrassata, rubicunda; petiolo vix ullo, eum glandula supraaxillari longe supra folii insertionem in ramo distincta : racemo in ramulis novellis terminali ; rachi valida, haud longa, compresse angulata; floribus maximis, speciosissimis, sparsis; pedicellis 1-floris, teretibus, crassis, longius- culis, imo foliolis seu bracteis majusculis 2 v. 3 obtuse ovatis concavis æqualibus patentibus persistentibus munitis; limbo calycino in alabastro maximo, ovato, integre clauso, demum in lobos 2 сопсауов parallele nervosos rupto; petalis 4, magnis, albis, 3plo longioribus, ovato-oblongis, subacutis, expansis: staminibus numerosissimis, pluriseriatis, imo breviter monadelphis, expansis, infra niveis, apice saturate rubris ; disco styloque ut in char. gen., intus breviter elevato: ovario infero, turbinato, vix angulato, 4-loculari, vertice intra discum concavo ; ovulis plurimis in quoque loculo suspensis: drupa magna, obpyriformi, vix 4-gona, apice sensim constricta et calyce coronata; pericarpio crasso, nitido, viridi-brunnescente ; meso- carpio spongiose carnoso; endocarpio coriaceo, fibris lignosis tecto, abortu uni- loculari et monospermo, dissepimentis hine in chordam crassam contractis et ad parietem adpressis ; semine magno, oblongo, obtuse 4-gono ; testa tenuiter coriacea, molliuseula, albo-incarnata; nucleo albo, sapore astringente, structura forsan aliorum. Іп insul. Otaheite : v. pl. s. in hb. Mus. Brit. Tahiti (Solander); fruct. non vidi. This species (here shown in Plate XL, taken from Parkinson's drawing) was col- lected by Solander during his three months’ residence in Otaheite, when he accom- panied Capt. Cook on his first voyage to the Pacific in 1769, to observe the transit of Venus; he found it also in the adjacent islands of Huabeine, Ulaietea, and Otaha. He describes it as a magnificent tree, 50 feet high, with a broad expanded head; its copious very large shining leaves, its splendid flowers, measuring 8 in. in diameter, interspersed with large bright pendent fruits, altogether form objects unsurpassed in beauty, and superb ornaments to the forests in the interior of the islands. Ву the natives it is called Аиа (Æhutu). Dryander, from Solander’s manuscript, regarded it as belonging to the Butonica of Rumph; but subsequently (in * Hort. Kew.’) he considered it identical with the Barringtonia speciosa of Forster, collected in Capt. Cook's second voyage, a plant evidently only known to him from Forster’s drawings Solander’s plant, though offering a general similarity to the latter, differs from it in its | * Primitie Flore insularum Oceani Pacifici, sive Cat. plant. in Otaheite, Eimeo, Otaha, Huaheine et Ulaietea, | 1769, diebus 13 Apr. ай 9 Aug. collectarum ; Solandri manuscript in Libr. Banks. Musso Britannico,—a work of | great value, never published, and but little known. — Parkinson's coloured drawings of the same plants, also unpub- lished, were made to illustrate that work. a аб Ал а еы ыа ы ыы MR. J. MIERS ON THE ВАККІМСТОМІАСЕ Ж. 61 much larger leaves, its flowers of greater size, on a longer raceme and on stouter pedicels, bracteated in the middle, larger calycine lobes, the presence of a large peculiar gland above the insertion of each leaf, and more especially by the different structure of its fruit and seeds. This last consideration places the two plants in distinet genera. Parkinson's exquisite coloured drawing gives an excellent representation of Solander's plant, but shows only the upper leaves, crowded on the summit of the branches as they appeared in nature ; but Solander's dried specimens show some of the lower leaves ; these аге 14-24 in. long, 83—12 in. broad, quite round at the apex, cuneated below into a broadish obsolete petiole, or expansion of the midrib. Solander states that an inch or less above the insertion of each leaf а reddish obtusely conical gland is observed on the branchlets. The rachis of the terminal raceme is stout; the pedicel of each flower, 6 in. long, 1 in. thick, bears in its middle а leaf-like sessile bract 21 in. long, 1 in. broad; the adnate portion of the calyx is $ in.long; the upper portion splits into two large concave lobes, with parallel nervures 11 in. long, 4 in. broad; the petals are 3} in. long, 1j in. broad ; the stamens, spreading in all directions, are agglutinated at the base into a mona- delphous tube 4 lines long, 8 lin. in diam., seated upon the outer margin of the disk, the slender filaments being 5 in. long; these, with the petals attached, soon fall off together ; the subulate persistent style is longer than them. Тһе fruit is 4 in. long, 2} in. broad below the middle, is crowned by the persistent calyx, disk, and style. Тһе thick peri- carp encloses a single seed, covered by a thin soft coriaceous testa, the nucleus, homo- geneous in texture, being white, oval, obsoletely 4-angled, the size of a pheasant's egg ; this is not edible, owing to its astringent flavour. In a note by the editor of * Cook's Voyages' (p. 157), it is stated that the plant is used by the natives of Otaheite to catch fish by intoxicating them. 2. ÀGASTA ASIATICA, nob.: Mammea Asiatica, Linn. Sp. Pl. i. p. 501 (1753); Osb. Itin. (1752), Voy. China, Engl. edit. 2, p. 62 (1771): Stravadium macrophyllum, Bl. in Van Houtte, Fl. Serr. vii. p. 24: Barringtonia macrophylla, Miq. Flor. Ind. Ned. i. p. 491: Barringtonia speciosa, Linn. fil Suppl. p. 312 (1781); Gaud. in Freyc. Voy. p. 482 (non Forster) ; Guillemin (in parte), in ** Zephyritis Taitensis,” Ann. Se. Nat. 2nd ser. vol. vii. p. 358 (1837); Paxton (non Forst.), Mag. Bot. x. р. 241 (cum icon. color.) ; Blume (non Forst. in Van Houtte, Flor. Serr. vol. iv. р. 409, cum icon. color. (ex icon. Paxt. clept.); Benth. (non Forst.), Lond. Journ. Bot. ii. 221; Flor. Austr. iii. p. 221; Seem. (non Forst.), Flor. Viti. i. p. 82; Lindl. Veg. Kingd. (ex Paxton), p. 755, fig. 503 * : arbor grandis, late frondosa, ramulis pendulis, crassis, striatis, rufo-brunneis, pruinosis, fistulosis : foliis maximis, con- gestis, elongato-oblongis, apice late ovatis, rotundatis vel obtusioribus, infra medium sensim angustatis, acutis, vel sepius circa petiolum rotundato-truncatis, integer- rimis vel subrepandis, chartaceis aut subcoriaceis, supra nitidis, pallide viridibus, nervis divaricatis arcuatim nexis, venis transversis reticulatis, costa plana, imo sensim incrassata, subtus pallidioribus, flavescenti-brunnescentibus, opacis, costa nervisque rubidulis, prominentibus; petiolo brevissimo, lato, supra plano, limbo SECOND SERIES.—BOTANY, VOL. I. K 62 MR. J. MIERS ОХ THE BARRINGTONIACE.E. 60plo breviore: racemo terminali, spieatim plurifloro, rachi crassa suberecta; floribus alternis, speciosis ; pedicellis validis, longiusculis, apice crassioribus ; calycis limbo maximo, primum oblonge globoso, integre clauso, parallele nervoso, demum in lobos 2 oblongos сопсауов Ғавсов minute granulosos rupto; petalis 4, duplo longioribus, obtuse oblongis, albis; staminibus sanguineis, disco styloque ut in char. gen.; ovario infero, turbinato, late 4-gono, 4-loculari ovulis in quoque loculo 5—6, suspensis: fructu majusculo, pyramidato, acute 4-gono, calyce coronato, imo umbilicato et cordatim truncato, abortu 1-loeulari et monospermo; регісагріо erassissimo, spongioso; endocarpio extus fibris lignosis crassiusculis intricatis tecto; embryone mesopodo. In Malacca, Java, Australia, insulis Maris Pacifici et Mauritianis: v. s. in herb. Mus. Brit. Java (Horsfield), in ins. St. Helena cult. (Home); in hb. Hook. Sincapoor (Anderson), Australia (Hill), Fiji (Barclay 8427), Fiji (Seemann 148) in ins. Maurit. (Blackburn), in ins. Mohelle (Meller); in herb. Soc. Linn. Penang (Wail. Cat. 3632 А), Sincapore (Wall. Cat. 3632 в), Cam- bodia (Wall. Cat. 3632 т): v. fruct. in Mus. Brit. et Kew. Linnzeus never saw this splendid plant, but described it in 1753 from the excellent materials furnished by Osbeck, who first observed it in 1752 upon an island near Java. Osbeck says it is а good-sized tree, often procumbent, with wide-spreading branches, hanging over the maritime shores or in the estuaries of rivers. He relates that he had much difficulty in collecting а specimen, owing to the furious attack of ants, which in myriads live in its fistulose branches. Тһе younger Linn:eus, in 1781, five years after the date of Forster's *Genera, and after his retirement to Germany, published a full description, purporting to be that of the Barringtonia speciosa, Forst., as obtained from Forster, and giving as its synonyms very different plants, the Mammea asiatica of Osbeck, the Commersona of Sonnerat, and the Butonica of Rumphius (tab. 114). Copying Rumphius, he states it to be a lofty tree, at the same time that he took the description of the leaves and inflorescence from Osbeck's details. А singular coincidence here occurs: his long diagnosis is in many parts a copy, word for word, of Forster's MSS description of his Barringtonia speciosa existing in the Paris Museum, according to Guillemin ; but the diagnosis of Linnzus omits all mention of the peculiar nectary and the unusual features given by Forster of the fruit and seeds—-characters that serve above all others to distinguish the two speciesas belonging to distinct genera. As it now stands, Linnsus's | diagnosis may be taken as a fair account of Agasta asiatica ; but as a description of Bar- | ringtonia speciosa it must be absolutely rejected. Osbeck states that its leaves are obovate, and more than 1 foot long; Horsfield’s вресі- | mens agree in form, аге 15 in. long, 7 in. broad, on а petiole 3 lines long, 2 lines | broad. Тһе specimens from Penang, Singapore, Cambodia, and other places above quoted, all accord in nearly similar dimensions and in the same length of petiole; but | Blume states that in his plant the leaves are 10—90 іп. long, 5-81 in. broad, on dixe 3 5-65 lines long—probably a mistake. . We have, however, an admirable account of this plant, introduced into Salisbury in à living state under the name of Barringtonia speciosa. Тһе main leaves grew to a length of more than a foot ; but the upper leaves became gradually smaller. Тһе plant flourished MR. J. MIERS ON THE BARRINGTONIACE Ж. 63 thirteen years, growing to the height of 8 feet; and it became necessary to cut off its top for the length of a foot, which cutting, when planted in a pot, grew rapidly, and produced a terminal thyrsoid inflorescence 22 inches long, from which the drawing of one of the flowers of the natural size, together with its floral leaf, was published by Paxton in 1845. This terminal inflorescence was 18-24 in. long, with axils about 1 in. apart. Тһе floral leaf is figured as 9 in. long, 3 in. broad, almost sessile; in the axil of each leaf a single large pedicellated flower originates. The lower pedicels of the full-grown flower аге 2 in. long; these gradually diminish in size to 3 in., supporting the young buds: the pedicels are stout and erect, each bearing on its summit the inferior ovary, crowned by the limb of the calyx, divided into two concave lobes nearly 13 in. long. It has 4 white petals, 2-22 in. long, 11—11 in. broad, attached by their claws to the staminiferous tube ; the filaments, of a sanguineous purple, are 3-43 in. long ; the style is 4—5 in. long. Only one flower blossomed at one time, which fell off next morning, when it was followed by others in succession. This account was published in * Paxton's Magazine.’ Van Houtte in 1848 copied the two drawings of Paxton in his ‘ Flore des Serres’ without acknowledgment, leading us to suppose they were made from a plant cultivated in Belgium. He omitted all the interesting observations of Paxton, but gave a description of it, copied word for word from Rumphius’s account of his Bufonica, also without the least mention of the source from which it was taken. This plant was introduced into the island of St. Helena prior to 1780, where it flourished in the garden of the Governor, and probably still exists there, as we know it was living a few years ago. The fruits in our museums are without localities; but those of this species are recog- nizable from those of A. indica by their larger size, their more regular pyramidal quadrate form with rounded angles, and more truncated (not cordate) base. The largest specimens I have seen are 41 in. long, 31 in. broad at the sides; others are 3} in. long: the peri- carp is polished, and of a pale colour, is very thick and spongy ; the endocarp is coriaceous, and is covered outside by ligneous fibres, which extend from the style to the top of the pedicel; a large exalbuminous embryo fills the cell, is 22 in. long, 13 in. broad, and in a transverse section shows a vacant space between the exorhiza and neorhiza. Seemann states that these large fruits are used by the natives of Fiji in their games, and that they serve as floats for their fishing-nets. As in the preceding species, they are poisonous, and are used to catch fish by stupifying them. In Fiji these fruits are called Уши, in the Tonga islands Futu, in the Pacific islands Hutu, Hootu, in Amboyna and Java Huttwm. Seemann adds that this species in Fiji is a magnificent sea-side tree, and that it furnishes the material liku, of which the women’s dresses are made. The analysis of the flower and fruit of this species is shown in Plate XII. 3. AGASTA INDICA, nob.: Barringtonia speciosa, W. & А. (non Forst. nec Roxb.), Prodr. Fl. Penins. p. 333 (excl. syn.); Wight, Icon. tab. 547 (excl. fig. fruct.) ; Thwaites, Enum. p. 119; sime nomine in Hermann, Icon. ined. tab. 241: arbor, ramulis crassis, cicatricatis, 4-angulatis, pallide brunneis, fistulosis; foliis ob- longis, apice gradatim obtusis, infra medium angustioribus, imo anguste rotundato- truncatis, marginibus cartilagineis integris, chartaceis, supra lucentibus vel sub- K 2 64 MR. J. MIERS ON THE BARRINGTONIACEA. opacis, pallide viridibus, nervis patentim divaricatis, costa plana imo sensim incras- sata, subtus pallidioribus, opacis, nervis costaque prominentibus ; petiolo brevissimo, late suleato: racemo terminali, thyrsoideo, erecto, folio breviore; floribus majus- culis, speciosis; pedicellis longis, teretibus, divaricato-erectis, imo bractea foliosa acute oblonga caduca munitis; calycis limbo magno, primum oblongo et integre clauso, parallele nervoso, demum in lobos 2 сопсауов rupto; petalis 4, duplo longi- oribus, acute oblongis, concavis, suberectis; staminibus numerosissimis, ista exce- dentibus ; disco intus urceolatim expanso styloque ut in char. gen. ; ovario infero, · turbinato, sub-4-gono, 4-loculari, ovulis in quoque loculo 6, axi superne affixis: fructu majusculo, calyce coronato, subglobose 5-angulato, apice subito attenuato, angulis aeutissimis in lobis basalibus continuis, imo plus minusve profunde cordatis et umbilicatis ; ceteris ut in specie precedente. Іп India orientali et Ceylonia indi- gena; in insulis-Comorinis et in Mauritio forsan introducta : v. s. in hb. Soc. Linn. Madras ( Wall. Cat. 3632 с); v. fructus in Mus. Kew. Comorin (Kirk). А tree of moderate size, with wide-spreading branches, at the extremities of which the leaves are approximated at a distance of 3 lines; the leaves vary in size from 44—12 in. long, 21—7 in. broad, on petioles 1-2 lines long and broad: the rachis of the raceme is 34 in. long; the lower pedicels 24 in. long. the upper ones in bud gradually much shorter, the basal bract in the former 1} in. long, 5 lines broad ; the calycine lobes are 1 in. long; the petals 13 in. long, 1 in. broad, falling off attached to the staminiferous tube; the filaments are 2 in. long; ovary 5 lines long. The fruit figured by Wight does not belong toit; he confesses that he never saw it, but copied it from Gaertner's tab. 101, which belongs to a very different plant—Butonica Витрата, from the Dutch Asiatic settle- ments, and the fruit of which accords with Sonnerat's drawing of it. Тһе fruit of the Indian species may be recognized by comparison with Hermann's drawing of it; the largest specimens in the British Museum (locality not cited) are 44 in. long, 44 on the sides, 5$ in. across the angles, the cordate basal lobes extending 1 in. below the attach- ment upon the pedicel. Kirk's specimen from Mohille is similar in shape, but only half that size; that in Hermann's drawing is intermediate in size, and similar to others in the British-Museum collection. This handsome well-marked species is distinguished from the two preceding by its more oblong leaves, shorter racemes, with smaller flowers, shorter petals, and by the shape of its fruit. Тһе upper leaves in Wight's plate 547 agree in size and shape with those in Hermann's drawing; but the lower leaves are gradually much larger. The specimen of the fruit found by Dr. Kirk on one of the Comorin Islands was pro- bably floated by sea from the Indian coast, as Commerson and Sonnerat relate, and is known by the name of “ Bonnet rouge." The floral and carpological features are shown in Plate XII. 3. BUTONICA. lis genus was first described in 1744 by Rumphius', was acknowledged by Lamarck in 1788 *, and by Jussieu i in 1189°; vat the younger Linnzeus, in 1781, united it with the EIL Ms ш 223 Diet, Méthod. i. 591. > Gen. Pl. p. 326. MR. J. MIERS ON THE БАВБІХСТОХТАСЕ Ж. 65 Barringtonia of Forster'. His example was followed by DeCandolle in 1728?, and by all sueceeding botanists, which has greatly contributed to the confusion at present ex- isting. The genus differs from Agasta very conspicuously: its leaves are more char- taceous, and serrated on the margins (not coriaceous and entire); its inflorescence is an elongated pendulous many-flowered spicated raceme (not erect and thyrsoid with few, very large flowers); the flowers are smaller, on much shorter slender pedicels, with a ealyx always conspieuously smaller, globose, and splitting into 3, rarely 4 lobes (not long, oval-oblong, and rupturing into two large concave lobes). It consists of several arborescent species, growing inland in moist places, or in estuaries on the sea-shores. The leaves are generally large, more or less acute and cuneated, and the racemes very long, with handsome flowers. Тһе fruit, in structure, resembles that of Agasta, and, owing to its thick spongy pericarp, is often found floating at sea along the Malayan coasts; they are also often found upon the coasts with the seeds germinating within the pericarp, after the manner shown by Roxburgh °. Butontca, Rumph., Lam. et Juss. Barringtonia, Linn. fil. (in parte) et alior.; Stravadium (in parte), DC.; Commersona et Menichea, Sonnerat. Calycis айпай! limbus mediocris, in alabastro globosus, integre clausus, mucronulatus, parallele nervosus, demum in lobos 3 vel 4 ruptus, viridis, persistens. Petala 4, triplo longiora, oblonga, unguibus tubo staminigero adglutinata et cum illo caduca. Stamina numerosissima, pluriseriata, imo breviter monadelpha, iis Barringtonie similia. Discus epigynus, plane aut pulvinatim annularis, margine exteriore tubum staminigerum fulciens, interiore in osculum elevatum expansus. Stylus et stigma ut in Agasta. Ovarium inferum, turbinatum, 4-loculare ; ovu/a in quoque loculo 2 vel plura, ab axi vel summo suspensa, anatropa. Fructus majusculus, cylindricus, ovatus, pyramidatus aut obpyri- formis, obsolete vel acute 4-gonus, aut 6-sulcatus cum alis 6 crassis pendentibus, apice subattenuatus | et calyce coronatus, indehiscens: pericarpium nitidum, crassissimum, spongioso-carnosum ; endo-' _ carpium tenuiter coriaceum, extus fibris numerosis lignosis tectum, intus abortu 1-loculare, cum loculis marcidis 3, in chorda longitudinali prominente eo affixa signatum, monospermum: semen magnum, loculo conforme; festa submembranacea, endocarpio adhesa; embryo exalbuminosus, solidus, amygdalinus, sepe edulis, ovatus, chord: impressione longitudinaliter canaliculatus, interdum striolatus, mesopodus. Arbores Asiatice, et in insulis maris Pacifici indiyene, ramosissime, ramulis sepe fistulosis : folia plerumque majuscula, oblonga, sepius utrinque acuta, plus minusve serrata, brevissime petiolata : racemi lami- nales, longi, spicati; flores speciosi, breviuscule pedicellati : fructus majusculus, forma varia, nitidus, levis, sepius ruber. 1. Вотомтсл ALBA, Rumph. Amb. у. cap. 30, p. 181, tab. 116 (ubi planta valde diminuta) : Stravadia alba, Pers. Ench. i. p. 30: Stravadium album, DC. Prodr. iii. p. 289: Barringtonia alba, Bl. in У. Houtte, Fl. Serr. vii. 24; Mig. Fl. Ind. Ned. i. 485 ; Hask. Bot. Zeit. xxvii. p. 598: Barringtonia conoidea, Griffith, № обще, р. 656; Icon. РІ. Asiat. tab. 635. fig. 1 ad 18 (plant. et flor. analysis) et tab. 634 (іп parte quoad semen?): arbuscula, ramulis crassiusculis, rugosis, subsuleatis, fistulosis, * Linn. fil. Suppl. Sp. pl. p. 312. > Prodr. iii. р. 288. 3 Wight, Icon. i. tab. 152. 66 MR. J. MIERS ON THE BARRINGTONIACEJE. e foliorum lapsu cicatricatis: foliis majusculis, elongato-oblongis, apice sensim aut repentius acuminatis, infra medium subeuneatis, imo acutis vel circa petiolum anguste rotundato-truncatis, marginibus cartilagineis undulato-revolutis, crenato- serratis, chartaceis, supra viridibus, opacis, sub lente impresso-rugulosis, nervis tenuibus arcuatim nexis prominulis, reticulatis, subtus pallide brunnescentibus, minutissime granulatis, costa striolata imo sensim crassiore, nervisque pallidissimis prominulis; petiolo crasso, semitereti, limbo 12—20plo breviore: racemis elongatis, pendulis, minute velutinis, floribus speciosis subsparsis, pedicellis imo articulatis et 2-bracteolatis; calycis limbo subparvo, globoso, demum in lobos 3-4 rupto; petalis 4, fere triplo longioribus, oblongis, lateribus retroflexis, carnosulis, albis ; staminibus, disco styloque ut in char. gen.; ovario infero, subcylindrico, 8-sulcato, 4.-loculari, ovulis in quoque loculo 2 apice suspensis: fructu ovali, utrinque rotundato, obsolete 4-сопо, calyce coronato, abortu 1-loculari et monospermo ; pericarpio viridi, crassis- simo; endocarpio coriaceo, extus fibris lignosis tecto; testa tenui, cum raphe ramosa parallele nervosa; embryone mesopodo, solide eburneo, 4—8-suleato. Іп Malacca et ins. Archipel. Asiat. in fluviatilibus : v. s. in hb. Hook. Mergui (Griffith), Malacca (Maingay 760); in hb. Soc. Linn. Sincapore (Wall. Cat. 36328), Penang (Wall. Cat. 3684 в et 3634 D), in hort. Sundriban cult. (Wall. Cat. 3634 A). The above specimens agree well with the description by Rumphius of his Butonica alba, the drawing of which in his plate 116 is on a scale so reduced as to give little idea of its character, all the parts being diminished to one fifth of their natural size. Griffith’s plant quite accords with the description; but in his plate the leaves and flowers are reduced to half their natural size; his manuscript details are very good and ample, and the analysis of the flower in its natural size is quite complete; his figs. 18—17 show sec- tions of the ovary, much magnified. Rumphius says Из. leaves are 18 in. long, 5-6 in. broad, with obsoletely crenulate margins, pallidly fuscous below, with raised white nerves; racemes 2—3 feet long; calyx divided into lobes; white petals, white stamens, red- dish at base; fruit pome-shaped, 3 in. long, 2 in. broad, deep green colour, marked by areolar impressions; nucleus white, not edible. 'The above diagnosis is framed upon Griffith's specimen and drawing; here the leaves are 102—121 in. long, 4-5 in. broad, on thick petioles 3-4 lines long (shown half size in his drawing); raceme 1-2 feet long, slender pedicels 4-6 lines long ; bracts nearly 1 in. long, very caducous; calyx globular in the bud, and 4 lines in diam., splitting at first into 2, then into 4 lobes; petals 10 lines long, 5 lines broad; the oval fruit is 3 in. long, 2 in. broad, with a pericarp 2-3 lines thick; seed oblong, tapering to its suspended summit, 15 lines long, 10lines broad. This seed appears represented in some of the figures of Griffith's plate 634, there being no room for it in his plate 635 (see under Careya spherica). In the memorandum attached to his specimen, he says the solitary seed is suspended, with a very thick coriaceous integument, is sulcate outside; its embryo is homogeneous, agreeing with the figures above mentioned. Its analysis is shown in Plate XIII. JTONICA RACEMOSA, Juss. Gen. 326 (excl. syn.) : Eugenia racemosa, Linn. Sp. РІ. 673; n так 191 me syn); Lam. Dict. iii. 197 (excl. syn.): Jambos sylvestris Saam- MR. J. MIERS ON THE ВАВБВІХСТОМТАСЕ Ж. 67 stravadi, Rheede, Hort. Mal. iv. р. 11, tab. 6; Willd. Sp. РІ. ii. 966 (excl. syn.); sine nomine, Hermann, icon. ined. tab. 212, 213 & 239: Barringtonia racemosa, Bl. (non Gaud.) in DC. Prodr. Ші. 288 (excl. syn.) ; Roxb. Fl. Ind. iii. 634 (exclus. syn. in parte); W. & A. Prodr. Fl. Ind. i. 133 (excl. syn.); Wight, Illustr. p. 19; Wight, Icon. tab. 152 (excl. fruct. fig.); Hook. Bot. Mag. tab. 3831; Wall. Cat. 3684 р; Thwaites, Enum. Pl. Zeyl. (excl. var. 8), p. 119: arbor alta, ramosa, trunco recto ; ramis crassis, cylindricis, e foliorum lapsu crebre cicatricatis; ramulis tenuibus pendentibus : foliis lanceolato-oblongis, infra medium ad basin gradatim angustatis, apice sensim longe acuminatis, marginibus subrevolutis, crenulato-serratis, tenuiter chartaceis, supra viridibus, nervis subpateutibus arcuatim nexis prominulis, venis transversim reticulatis, costa plana, imo latiore, subtus flavide opacis, nervis promi- . nentibus ; petiolo latiusculo, supra sulcato, late marginato, rubro, limbo 60plo bre- viore : racemo terminali, pendulo, glaberrimo, rachi valida, longa, spicatim multiflora ; floribus speciosis, breviter pedicellatis; pedicellis imo bractea minuta caduca munitis ; calyce globoso, demum in lobos 3-4 rupto; petalis acute obovatis triplo longioribus, marginibus retroflexis, flavidis; staminibus rubris, disco, styloque ut in char. gen. ; ovario infero, turbinato, acute 4-gono, 2-3-loculari, ovulis plurimis in quoque loculo, ab axi radiantibus: drupa majuscula, obverse obovata, apice sensim coaretata, ubi 4-suleata et calyce coronata; pericarpio crassissimo, extus viridi; mesocarpio albido, demum brunnescente, spongioso; endocarpio fibris lignosis tecto, abortu 1-loculari et monospermo; semine magno, oblongo, utrinque acuto, appenso, homogeneo, ab utraque extremitate germinante. In Indie peninsula: v. pl. s. in hb. Soc. Linn. Sundriban (Wall. Cat. 3634 A) ; hb. Finlayson (Wall. Cat. 3634 D). This is described by Rheede as a tree of vast size, growing in moist woods, and by Roxburgh as a lofty tree, with a head of many spreading branches; its trunk gives a dense wood of a yellowish white colour; its leaves are 3-15 in. long, 13-4 in. broad, on a broad petiole 3—4 lines long. Тһе rachis of the raceme is 2 feet long, and expands to a width of 11-2 in., with flowers 4—2 in. apart, on pedicels 3-6 lines long; calycine lobes after expansion 4 lines long; енің 9 lines long, 4 lines broad, white; filaments 13 in. long, white, agglutinated at base into a tube 3 lines long. The drawing of Her- mann agrees with that of Rheede as to the size and shape of the fruit, which is 3 in. long, 21 in. broad; that in Roxburgh's figure is similar in shape, 2$ in. long, 2 in. broad below the middle, and shows the seed germinating within the регісагр, throwing out a young stem from the extremity near the pedicel, and a rootlet from the apex of the fruit— а development analogous to that described by Rumphius in Butonica Rumphiana; it Shows that the seed must be suspended, and that the plumular extremity of the neorhiza points to the base of the fruit. A very good representation of the plant is given by Sir William Hooker, drawn from a specimen cultivated in Liverpool, and received from Bom- bay. In three years’ time, when 8 feet high, it yielded flowers in a raceme 28 in. long. Linnzeus first described the species in his Ceylon Flora, on a plant not seen by him, upon the drawings of Hermann’s Ceylon plants, and again in his Sp. Plant. по. 5, where he considered it to be identical with Rheede's Malabar Saamstravadi, tab. 7. 68 MR. J. MIERS ON THE BARRINGTONIACE./E. 3. Вгтохтса RuMPHIANA, nob.: Butonica, Rumph. Amb. v. сар. 29, р. 179, tab. 114 Lam. Encyl. i. 521 (excl. synon.): Barringtonia speciosa, Gaertn. (non Forst.), De Fruct. ii. p. 96, tab. 101 (excl. synon.); Roxb. Fl. Ind. iii. 636; ВІ. (іп parte), Van Houtte, Fl. Serr. vii. p. 23: Barringtonia racemosa auct. (in parte): Com- mersona, Sonnerat, Voy. Guin. p. 14, tab. 8 et 9: Mitraria Commersoni, Gmelin, Syst. p. 799: Fructus peregrinus tetragonus, Clus. Exot. lib. 2. cap. 5: Stravadium rubrum, DC. (in parte), Prodr. iii. 289: arbor, ramulis cinereis, apice crebre foliosis; foliis majusculis, elongato-oblongis, apice sensim angustioribus et breviter acuminatis, infra medium gradatim angustioribus, et imo rotundiuscule truncatis, margine ser- rulatis, supra profunde viridibus, nervis tenuibus costaque subimmersis, subtus pallidioribus, costa nervisque valde prominentibus; petiolo semitereti, limbo 70plo breviore: racemo elongato, pendulo, pruinoso; floribus speciosis, sparsis; pedicellis subbrevibus, bractea foliosa valde caduca munitis; calyce mediocri, globoso, demum in lobos 2, dein in 4 rupto; petalis 4, duplo longioribus, oblongis, concavis, albis; staminibus, disco styloque ut in char. gen.; ovario turbinato, 4-loculari, ovulis in quoque loculo 2 appensis: drupa mitreeformi, infra medium obtuse 4-angulata, superne sensim angustata, imo truncata et umbilicata, extus rubra, aut sicca fla- vescente, nitida, lobis calycinis immutatis coronata. In insulis Archip. Asiat. et in Siam; v. s. pl. in hb. Hook. Siam (Schomburgk); v. fr. s. іп Mus. Brit. This is described by Rumphius as а lofty tree growing near the sea-shore, throwing out many spreading branches after the manner of Terminalia Сарра, with leaves 16-18 in. long, 41-6 in. broad, on a petiole 12 line long and half that breadth, all much reduced in his drawing. In Sonnerat's drawing the flowers expanded, 2 inches in diameter, are 3 lines apart, on a stout rachis 3 in. long, with bracts 1i in. long, 1 in. broad ; pedicels 3 lines long; lengthening with the growth of the fruits, when they bend backwards; calyx in bud globose, 6 lines in diameter, splitting into four persistent lobes; petals 8 lines long; stamens 1 inch long. The drawing of Rumphius agrees with that of Sonnerat as to the size n shape of the fruit, and is identical with that of the Barringtonia speciosa of Gaertner (non Forst.) in his plate 101; it 15 2} in. long, 21 in. broad, with quadrate sides, 4 obtuse angles, gra- | dually contracting into a narrow neck at the summit, where it is crowned by the per- - | sistent calyx. Rumphius describes the seed as germinating within the fruit as it lies 00 the sea-shore, when a new shoot or stemlet first protrudes from the end, close to the pedicel, while a new rootlet is emitted from the apex, just as Roxburgh figures it in В. race- | mosa, He also mentions a kindred species from Celebes, where the fruit is the size of à | child’s head. Sonnerat, who gives a good account of this plant, which he found at Pulo Penang, says that its fruit is frequently found at sea, floating between the islands. _ Schomburgk’s specimens are referred to this species, because they agree in the size and | shape of its leaves, which are 7-103 in. long, 3-4} in. broad, on a petiole 1-14 line long; | the raceme is about 9 in. long, with flowers about 4 in. apart; the globular calyx in bud | is 5 lines in diam., splits first into two, then into 4 pom the petals are 9 lines long, the E ovary 4-celled. For the structural characters of this species see Plate XIII. MR. J. MIERS ON THE BARRINGTONIACEA. 69 4. BuTONICA TERRESTRIS, Rumph. Amb. iii. lib. 5. cap. 30. р. 181, tab. 115 (pl. ad 4 dimens. reduct.): Barringtonia elongata, Korth. in Kruidk. Arch. i. p. 206 (nomen solum) : Barringtonia racemosa, Bl. (non Juss.), in Van Houtte's Flora, vii. p. 28; Miq. Fl. Ned. Ind. i. p. 486: Barringtonia rubra, Miq. l. c. p.487: Stravadium rubrum, DC. Prodr. iii. 289 (excl. syn.): arbor demissa, ramulis crassiusculis, fistu- losis, rufo-brunneis, pruinosis: foliis approximatis, majusculis, elongato-oblongis, apice acute attenuatis, imo spathulato-cuneatis, angustatis, et in petiolo decurren- tibus, marginibus subrepandis et supra medium sinuato-serratis, dentibus obtusulis вгере aciculato-mucronatis, tenuiter chartaceis, supra pallide viridibus, opacis, nervis adscendentibus utrinque circa 20 tenuibus prominulis et marginem versus arcuatim nexis, subtus flavide pallidioribus, opacis, nervis costaque lata striolata prominenti- bus; petiolo subtenui, limbo 16plo breviore: racemo terminali, pendulo, folio vix longiore; rachi subtenui, striata; floribus alternis, subremotis, breviter pedicellatis ; calycis limbo mediocri, globoso, demum in lobos 3 ovatos concavos rupto; petalis 4, duplo longioribus, obovatis, rubris; staminibus, disco styloque ut in char. gen.; ovario turbinato, sub-4-gono, 4-loculari, ovulis in quoque loculo 2 suspensis: fructu oblongo, utrinque obtuso, ealyce coronato, costato-4-gono, rugoso, viridi, demum flavide brunneo, abortu 1-осшагі et monospermo; seminis embryone mesopodo, rotunde oblongo, obsolete 4-gono. In Java orientali indigena et in hort. Bogor. cult.: v. s. in hb. Hook. Banca (Horsfield), Patjetan (Horsfield 209). The specimens above quoted accord well with Rumphius's Butonica terrestris rubra in the size and shape of the leaves as well as in the inflorescence, agreeing with the dimensions given in the text; but on his plate 115 the plant is reduced to a quarter of its proper size, which makes it difficult of recognition; we cannot hesitate, however, in considering them specifically identical. Rumphius describes it as a low tree, growing in gravelly soils, in woods remote from the sea. According to his account, the leaves are narrower and more flaccid than in the other plants described by him; they are acute at both extremities, obsoletely serrulate, 12—18 in. long, 3 in. broad, and petiolated ; the raceme is 2 feet long, with somewhat remote flowers; the calyx is divided into 2 or 3 lobes; petals and stamens red; the fruit, crowned by the calyx, is 3 in. long, 15 іп. broad. In Horsfield’s specimens the leaves are 11} in. long, 2% in. broad, on petioles 8 lines long; the raceme is above a foot long, with flowers ? in. apart, on pedicels 1 line long; the calycine lobes are 4 lines long; the petals 8 lines long, 6 lines broad, aggluti- nated by their claws to the staminiferous tube. The analysis of the flower, and a drawing of the fruit in its natural size, are shown in Plate XIV. figs. 4-9. The Barringtonia racemosa of Blume (non Juss.) is considered by Miquel to be the same as Horsfield’s plant from Patjetan and Banca, which is here regarded as identical with Rumphius's plant from Amboyna, in the same region. Blume's species, В. race- mosa, in Van Houtte's Flora, is a heterogeneous mixture of plants from all quarters. There can therefore be no hesitation in placing under this species the Java plants enumerated by Blume and Miquel under the name of В. racemosa, that species being confined to the Indian peninsula and Ceylon. SECOND SERIES.— BOTANY, VOL. I. L 70 MR. J. MIERS ON THE ВАВВГУСТОМТАСЕЖ. 5. BuTONICA RUBRA, nob.: Butonica Tsjeria Saamstravadi, Rheede, Hort. Malab. part iv. р. 15, tab. 7: Eugenia acutangula, Linn. (in parte), Sp. Pl. 1. 471: Stravadium rubrum, DC. (in parte), Prodr. iii. 989: Stravadia rubra, Pers. (in parte), Ench. i. 30: Barringtonia rubra, Bl. in Van Houtte, Fl. Serr. vii. 28: arbor alta, ramulis pallide brunneis, lignosis, angulatis, rugulosis: foliis approximatis, ellipticis, apice in acumen acutum subito constrictis, infra medium spathulato-angustatis, circa petiolum repente rotundato-truncatis, obsolete et obtuse serratis, chartaceis, supra seepe pallidissimis, albidule opacissimis, aut pallide viridibus, nervis patenti-divari- catis venisque transversim reticulatis vix prominulis, costa plana, tenuiter carinata, et imo subdilatata, subtus pallidis, opacis, costa striolata nervisque prominulis; petiolo latiusculo, supra plano, subtus convexo et striolato, limbo 20plo breviore: racemo terminali, longissimo, pendulo, remote plurifloro ; floribus tenuiter pedi- cellatis ; calycis limbo mediocri, globoso, demum in lobos 8-4 oblongos subcoriaceos parallele nervosos opace granulosos rupto; petalis 4, duplo vel triplo longioribus, oblongis, purpureis ; staminibus, disco styloque ut in char. gen. ; ovario infero, turbi- nato, 4-loculari, ovulis in quoque loculo 2 suspensis: fructu oblongo, utrinque rotun- dato, calyce coronato, sub-4-gono, angulis rotundatis, abortu 1-loculari et mono- spermo; pericarpio crassissimo; semine cylindrico. Іп penins. Ірі: v. s. in hb. Hook. Quilon (Wight 1064), Concan (Horter); in hb. Soc. Linn. hort. Cale. cult. (Wall. Cat. 3634 с). This is a second exclusively Indian species, sufficiently well described by Rheede, which has been strangely confounded with Rumphius’s B. terrestris from Amboyna and Java, and also with Stravadiwm acutangulum, by all botanists. The mistake originated with Linnzeus, who confounded Rumphius’s plant with Hermann’s, and was unacquainted with any of the plants mentioned by him as synonyms. It is, however, a well-marked species. Rheede describes it as a tall tree; but his drawing of a fruit-bearing branch, being reduced to half its due proportions, renders it difficult of recognition. It differs, however, from B. terrestris in its much shorter and much broader leaves; and its fruit, though somewhat similar, is longer and narrower. It differs from В. racemosa for the same reasons. It bears no resemblance whatever to Stravadiwm acutangulum. Its leaves, about 2 in. apart, аге 63—83 in. long, 31—4 in. broad, on petioles 4-5 lines long; the slender raceme is 28-86 in. long; the slender pedicels, 1 in. apart, are 3 lines long; the calycine limb is globular and entire in the bud, 5 lines in diameter, soon splitting into 3 equal lobes ; the petals are 9 lines long; the outer series of filaments are 11 lines long, the inner series being gradually shorter. The fruit, as shown іп Rheede’s drawing. when restored to its proper size, would be 32 in. long, 11 in. broad, of an oblong shape. rounded at both ends, quadrately cylindrical, with obtuse angles, with a thick pericarp enclosing a cylindrical nucleus, said to be of a sweetish taste at first, but afterwards having an unpleasant bitter flavour. Тһе fruit in its natural dimensions is shown in Plate XIV. fig. 2, the flower in fig. 1. 6. Вотом1сА АТАТА, nob. : Barringtonia alata, Wall. Cat. (non Miq.); Griffith, Notule, 1 iv. p. 636; Icon. Pl. Asiat. tab. 636. figs. 1-6 (in errore sub Barringtonia conoidea) : 3 MR. J. MIERS ON THE BARRINGTONIACEÆ. 71 fruticosa, ramulis brunneis, pallide glaucis, interrupte striatis, apice confertim foli- osis; foliis oblongis, apice in acumen breve obtusum constrictis, basin versus angus- tatis et subito rotundiusculis, obsolete serratis, dentibus obtusissimis, chartaceis, supra profunde viridibus, opacis, nervis vix prominulis, subtus pallidioribus, opacis, costa nervis venisque reticulatis prominulis ; petiolo latiusculo, limbo 40plo breviore : racemis axillaribus, subspicatim erectis, folio dimidio vel bis triente brevioribus, sub- paucifloris; pedicellis tenuiter teretibus, imo articulatis et minute bibracteolatis ; calycis limbo mediocri, primum globoso-ovato, apice mucronulato, demum in lobos 2, dein in 4, rupto ; petalis 4, oblongo-lanceolatis, carnosulis, albide carneis ; staminibus, disco styloque ut in char. gen. ; ovario infero, conico-cylindrico, 6-sulcato, imo um- bilieato et 6-lobo, 4-loculari, ovulis in quoque loculo 2 suspensis: drupa majuscula, conice elongata, imo profunde umbilicata, apice 6-sulcata et calyce coronata, basin versus in lobos 6 crasse alatos pendentes expansa, 1-loculari, abortu 1-sperma ; peri- carpio sicco, fibroso ; semine majusculo, ovato, 6-suleato, ab apice pendulo; embryone conformi, homogeneo, neorhiza interna conspicua. In Malacca: v. s. in herb. Hook. Mergui (Griffith) ; in hb. Soc. Linn. Moulmein (Wall. Cat. 3633); v. fruct. in Mus. Soe. Linn. ( Wallich). We have no published aecount or manuscript details of this remarkable species. Griffith's analysis of its fruits, which agrees with all I found in Wallich’s specimen (as shown here in Plate XIV.), is figured in his work in plate 636 в, figs. 1 and 6, through the mistake of the editor of the * Хобе, under the name of Barringtonia conoidea, while the fruit of the latter is not represented by him, though it is well described in a manu- script note attached by him to his specimen of the plant. The leaves are 7-10 in. long, 21-31 in. broad, on a rather wide flat petiole 13-2 lines long, and they have about 11 pairs of curving subascending nerves; the racemes are 3—4 in. long, with a slender, erect, compressed, striolated, brownish-opaque rachis, bearing alternate flowers 3 in. apart; the slender subnutant pedicels are 9 lines long; the calycine lobes, after bursting, are 41 lines long, 2-8 lines broad, pruinosely opaque, concave ; petals pruinose outside; the fruit, including the wings, which descend $ in. below the top of the pedicel, is 2} in. long, 2 in. near the apex, gradually widening to a breadth of 2 in. across the six compressed subcoriaceous wings; pericarp 14 line thick, with an oval cell some- what narrow at the apex, the axis and dissepiments being pressed together on one side ; seed oblong, 6-suleated, suspended by a funicle, around which the abortive ovules are seen; the hard homogeneous nucleus is lj in. long, 1 in. in diameter, the internal neorhiza having a diameter of half that size, and is partially free from the exorhiza, showing a sensible interval between them. Ап analysis of Wallich's specimen of the fruit is shown in Plate XIV. figs. 13, 14. 7. Butontca INCLYTA, nob.: Barringtonia racemosa, Griffith (non ВІ. nec alior.), Notule, p. 659; Icon. РІ. Asiat. tab. 636. 2 (non 1). fig. 1 ad 6: ramulis sub- tenuibus, apice, erebre foliosis: foliis mediocribus, obverse obovato-oblongis, apice . valde obtusis aut rotundatis, sæpe emarginatis, infra medium subcuneatim angus- | L4 72 MR. J. MIERS ON THE BARRINGTONIACE.JE. tatis, circa petiolum subito anguste rotundiusculis, marginibus subintegris, valde revolutis et subretuse undulatis, subcoriaceis, supra pallide viridibus, valde opacis, ruguloso-granulatis, nervis adscendenti-divaricatis, semiimmersis, reticulatis, subtus rufescenti-pallidis, opacis, nervis venis costaque plana et striolata prominulis; petiolo brevissimo, aut vix ullo: racemo axillari; floribus ignotis: fructu majusculo, obpyriformi-oblongo, 8-sulcato, calyce coronato; регісагріо crassissimo, ligneo- fibroso, abortu 1-loculari et monospermo; semine magno, ab apice suspenso, 10-sul- cato ; testa subcoriacea, cum racheos ramis longitudinalibus in suleos immersis ; em- bryone conformi, mesopodo, apice in germinatione radicem pullulante, imo caulem novum emittente (sec. cl. Griffith). In Malacca: v. pl. s. т hb. Hook. ( Griffith) sine flore. A species well distinguished by the size and shape of its leaves, and by the larger size of its fruit, which is differently constructed from that of Butonica racemosa, to which Griffith referred it. Не gives no description whatever of his plant, but he figures the fruit. In his ‘ Мой’ he states that he found 4 species of Barringtonia in Malacca :— Ist, this plant, of which he sent home a specimen. 2nd, his В. conoidea, well repre- sented separately in the plant, its inflorescence, and also in its fruit, in his plate 636, and which I have referred to В. alata. 3rd, B. alata, of which no account is given, but its fruit is well shown in his plate 635, and which his editor mistook for that of В. conoidea, confounding together Griffith’s explanations of the two; both the fruit and specimens of the plant are manifestly identical with Wallich's plants of the same. 4th, В. cylindrostachys, which I have referred to the genus Doromma. Its approximated leaves are 41—61 in. long, 2-21 in. broad, on broadish petioles, 11-2 lines long; a bare raceme is seen in the specimen, the rachis of which is 6 in. long, with cicatrices of the fallen flowers + in. apart. The fruit in Griffith's drawing is large, 33 in. long, 21 in. in diam., rounded at the base, tapering a little above the middle, and crowned by the persistent calyx; the fibrous pericarp is 6 lines thick ; the suspended seed as large as a fowl’s egg, showing at its summit the abortive ovules, and the axis and abortive dissepiments pressed to one side, appearing like a thick longitudinal raphe; but the true raphe is seen in ten longitudinal branches imbedded in the thickish testa, which are opposite to as many furrows in the large embryo. This is drawn in its full size, showing the commencement of its germination by the protrusion of a small nipple in the summit, which afterwards expands into a root, while the bottom of the exorhiza splits to allow the exit of the extremity of the neorhiza, afterwards becoming the new stem, charged with scales; and upon this structure and on the mode of its germination, Griffith offers the remarks to which I have before alluded (р. 50). Тһе fruit of this species is shown in Plate XIV. fig. 19.). . BUTONICA ROSATA, nob.: Menichea rosata, Sonnerat, Voy. Guin. (1776), p. 133, tab. 92, 93: arbor ramis crassis: foliis oblongo-lanceolatis, apice sensim acute acumi- natis, imo longe cuneatis, obsolete serrulatis, petiolo brevi: racemis terminalibus vel е trunco enatis, nutantibus, folio subzequilongis ; floribus alternis, pedicellatis ; calyce demum in lobos 3 acutos rupto ; petalis 4, albis, concavis, apice rotundatis; stami- nibus, disco styloque ut in char. gen. ; ovario infero, turbinato, acute 4-gono, 4-locu- MR. J. MIERS ON THE BARRINGTONIACEA. 13 lari: fructu oblongo-ovato, utrinque angustiore, 4-gono, obtuse angulato, breviter stipitato, calyce coronato; pericarpio roseo, carnosulo, abortu monospermo ; semine ovali, apice acutiore, funieulo brevi sub apicem suspenso; testa parallele pluri- nervoso; embryone valde oleaginoso, eduli. In insulis Philippinis et Formosa: v. pl. sic. іп hb. Mus. Brit. Formosa (Oldham 115). This is described by Sonnerat as a tree growing in moist places at Sambouangue and elsewhere. Oldham's specimen from the adjoining island of Formosa agrees well with Sonnerat's drawing in plate 92, where the plant is represented in one third of its natural size, judging from the comparative dimensions of the calyx there shown—and in his plate ` 93, where it and the fruit are drawn in their proper dimensions. The leaves in Oldham's plant agree in shape and size with those of Sonnerat, restored to their full size: these are 83-9 in. long, 23-3 in. broad, on shortish petioles. Тһе raceme, restored to its due size, is 9 inches long, bearing about 12 alternate flowers, on slender pedicels 6 lines long ; the lobes of the calyx are 4 lines long, 21 lines broad. Тһе fruit, shortly stipitate, is 31 in. long, 2 in. in diameter, with a rose-coloured subfleshy pericarp; its single seed is 21 in. long, 11 in. broad ; according to Sonnerat, it is very oleaginous and edible, forming one of the best fruits of the island, where it is called Jam rosata. The fruit is shown in Plate XIV. fig. 17. 9. BUTONICA INTERMEDIA, nob.: Barringtonia intermedia, Vieillard, Bull. Soc. Linn. Normand. vol. x. p. 4: Barringtonia racemosa, Seem. (non Bl.), Flor. Viti. p. 683: Eugenia racemosa, Forst. (non Linn. пес DeCand.), Prodr. Fl. Austr. 39; Vieillard l.c. p. 9: arbuscula: foliis summo ramulorum dense confertis, lanceolato-oblongis, apice acutis, infra medium spathulato-angustatis, ad basin arctatam truncato-rotun- datis, margine grosse crenatis vel obtuse serratis, conduplicatis, glabris, supra pallide viridibus, nervis subpatentibus paullo prominulis intra marginem arcuatim nexis, costa tenui, deorsum sensim incrassata, subtus pallidioribus, nervis cum aliis inter- mediis brevibus venisque valde reticulatis prominentibus, costa valida in petiolum brevissimum latum desinente: racemo terminali, pendulo, folio 3-4plo longiore; raehi subtenui, angulata, striata, pruinosa, basin versus bracteolata, mox sparse florifera; floribus mediocribus; pedicellis tenuibus, 4-gonis, nutantibus, imo brac- teola lineari-lanceolata caducissima munitis; calycis limbo primum globoso, mucro- nato, integre clauso, demum in lobos 3—4 inæqualiter rupto; petalis 4, crassiusculis, albido-violaceis; staminibus numerosissimis, imo monadelphis, interioribus вере anantheris; ovario infero, turbinato, 4-gono, 4-loculari, ovulis in quoque loculo 4, pendulis: fructu oblongo, utrinque subalatim 4-angulato ; pericarpio coriaceo, sub- tenui. In insulis Neo-Caledonia et Vitiensibus : v. 8. in hb. Hook. Масар (Vieillard 2239), Kanalak (De Planche 86,87); in hb. Mus. Brit. New Caledonia (Forster sub Eugenia racemosa), New Caled. (Anderson sub Е. racemosa), Fiji ins. Taviuni (Seemann 149). А tree 10-20 feet high, according to Vieillard, found in maritime places or in running streams on the sea-coast (Seemann). The leaves in all the above specimens resemble each other in their shape and venation, varying from 10 to 15 in. long, 24-34 in. broad, on 74 MR. J. MIERS ON THE BARRINGTONIACEÆ. ‘petioles 1-2 lines long; the racemes, in the specimens, аге 10-19 in. long; but Vieillard says they are often "p 21 to 3 feet in length: he also states they are sericeous ; but in his specimen, as well as in all the others, they are glabrous and pruinosely opaque; the slender pedicels, $ in. apart, are generally 3—1 in. long, shorter in the younger flowers; the globose calycine limb is 4} lines in diameter, and splits into 3 lobes 5 lines long; the petals are 1 in. long, 6 lines broad, with the lateral margins retroflected the ovary is 23 lines long, with 4 almost winged angles. Seemann's specimen of the fruit, apparently not ripe, accords with the description of УіеШага; it is oblong, acute at base, narrowed above, 2 in. long, 3 in. broad, crowned by calycine lobes, narrowly winged on its angles, and contains a single suspended seed with a mesopodal embryo. This fruit, according to Seemann, is poisonous, and used for stupifying fishes; it is called Vutu-ui-wai (Water-vutu). 10. BUTONICA PROCERA, nob.: Barringtonia excelsa, Benth. (non Bl.), Lond. Journ. Bot. ii. 221: arbor elata, ramulis in summo сгерге foliosis: foliis majusculis, elongato- oblongis, apice breviter aut sensim acuminatis, infra medium spathulato-angustatis, imo obtusis vel rotundato-truncatis, marginibus revolutis, sub medium integris, superne crenato-serratis, chartaceis, supra pallide viridibus, opacis, nervis conspicuis divaricatis intra marginem nexis, costa supra plana, versus imum sensim incrassata, subtus pallidioribus; petiolo vix ullo: racemo subterminali solitario (vel interdum 3 subaxillaribus congestim approximatis) ; rachi longissima, pendula, folio triplo lon- giore, glabra, acute angulata, striata; floribus inferioribus sparsis, superioribus crebrioribus; pedicellis brevissimis, crassis, ebracteatis, floribus hine subsessili- bus; calycis limbo primum globoso, integre clauso, dein in lobos 2 apice sspe bifidos coriaceos concavos pruinosos extus granulatos rupto; petalis 4, triplo longi- - oribus, oblongo-ovatis, albis, membranaceis, lateribus subretroflexis, unguibus tubo staminifero agglutinatis; staminibus, disco styloque ut in char. gen.; ovario infero, cylindrico, medio ventricoso, 8-costato, cinereo-pruinoso, 4-loculari, ovulis in quoque loculo 4 per paria axi pendulis: fructo ignoto. In insulis maris Pacifici: v. s. М herb. Hook. New Hebrides, in Tanna (Hinds) ; in hb. Mus. Brit. Tanna (Hinds). А very peculiar species, referred by Mr. Bentham to Barringtonia excelsa, Bl., which | is a very different plant. Hinds states that it is a large, lofty, handsome tree; it must not be confounded with specimens collected by Barclay at the same time in the same | island, which I have referred to В. samoénsis. Its leaves are 12-18 in. long, 81—51 mm. - broad, on petioles scarcely appreciable in length, being a short extension of the costa, | 2 lines broad; the raceme is single, and nearly terminal іп Hinds's specimen; but he 1 remarks, in a note, that he had seen three of these spicated racemes forming only a part 1 of а very magnificent cluster of flowers; the rachis in the specimen is as long as the : leaves, somewhat slender, bearing many spicated flowers, almost sessile, the lower ones | 1} in. apart, gradually closer upwards ; the calycine lobes are 4 lines long ; the petals | 12 lines long, 5 lines broad; the outer series of stamens somewhat exceed the length of © the petals ; the circular disk supports on its outer margin the staminiferous tube ; and its : MR J. MIERS ON THE BARRINGTONIACES. 75 inner margin rises іп a short membranaceous erect tube, leaving a deep hollow, radiately striated, between it and the style. 11. BuTONICA SAMOENSIS, nob.: Barringtonia samoénsis, A. Gray, Un.-St. Expl. Exped. р. 508; Walpers, Ann. iv. 852: Barringtonia excelsa, Gray (non. Bl. пес Benth.) loc. cit. p. 508: Barringtonia racemosa, Gaud. (non Bl.), Freye. Voy. p. 483, tab. 107 (excl. syn.) : Stravadium insigne, Bl. in Van Houtte, Flor. Serr. vii. p. 24, tab. 654, 655: Barringtonia insignis, Miq. in Flor. Ned. Ind. i. p. 488: Barringtonia acutangula, Bl. (non Roxb.), Bijdr. 1097, sec. Miq. loc. cit.: arbor, ramulis subcerassis, striatis : foliis subapproximatis, lanceolato- vel obovato-oblongis, apice acute attenuatis, imo sensim angustioribus et subobtusis, marginibus vix revolutis crenato-undulatis, sub- serratis, flaecidis, submembranaceis, supra viridibus, opacis, ad nervos suleatis, nervis tenuibus adscendentibus venisque rubellis reticulatis paullo prominulis, subtus pallidioribus, opacis, minute granulatis, costa nervisque prominulis ; petiolo semitereti, latiusculo, limbo 33plo breviore: racemo terminali, pendulo; rachi gracili, elongata; floribus speciosis, alternis, breviter pedicellatis; calycis limbo mediocri, primum integre globoso, demum in lobos 3—4 rupto; petalis 4, triplo lon- gioribus, obtuse oblongis, suberassis, marginibus membranaceis undulato-subrecur- vatis; staminibus, disco styloque ut in char. gen.; ovario infero, turbinato, 4-gono, 4-loculari, ovulis in quoque loculo 4, suspensis : fructu longiuscule oblongo, utrinque acutiore, lobis calycinis coronato, 4-zono, angulis costatis, abortu l-loculari et monospermo, semine suspenso, embryone mesopodo. Іп insulis Oceani Pacifici : t. 8. in hb. Hook. Port Resolution, Tanna (Barclay 3487); in hb. Mus. Brit. Tanna (Barclay). А species hitherto found only in the New Hebrides, Navigators', and the Ladrone or Mariana group of islands. It is said to be a tall handsome tree, growing in moist places ; it is well represented in Gaudichaud's drawing, the leaves of which approach those of Dr. Gray's var. В. Тһе plant is everywhere glabrous; the leaves are 8-15 in. (some- times 24 in.) long, 21-31 in. broad (in var. В 45 in. broad), on petioles З lines long and 1$ line broad, with about 18 pairs of subascending nerves, with others shorter and inter- mediate, all anastomosing ; the terminal raceme is 15-24 in. long, with a rather slender rachis, bearing alternate flowers 3 in. apart, which, when expanded, are 2 in. across ; pedicels 2—4 lines long; calycine limb at first globose, 3 lin. in diam., splits into 3-4 lobes ; petals 10 lin. long, 5 lin. broad; the fruit, as shown in Gaudichaud’s drawing, is 3 in. long, 17 in. broad, with 4 costate angles, decurrent on the pedicel. Gaudichaud’s description of the seed is, that it is rostellated and bilobed at the apex. This would be ‚ better interpreted by saying that, as in all the other species, the seed is suspended by a funicle, to which the abortive ovules are also attached, and which seems lobed at its apex, as Griffith shows in his Icones, pl. 635. fig. 8. Van Houtte’s drawing, tab. 654, to Which he gave the name of Stravadium insigne, Bl., does not appear to have been made from any cultivated specimen, as he would lead us to suppose; its originality may indeed be doubted, after what I have shown in regard to his drawing, tab. 409, under Barring- 76 MR. J. MIERS ON THE BARRINGTONIACEJE. tonia speciosa. From the resemblance of the former, in the form and size of the leaf and the size of the flowers, we may infer that it is a made-up drawing, with its details bor- rowed from that of Gaudichaud. Тһе analysis of the flower and the fruit are shown in Plate XIV. figs. 20-25. 12. BuroNICA EDULIS, nob.: Barringtonia edulis, Seem. Fl. Vit. p. 82: arborea: foliis elliptico-oblongis, apice sensim acutis, infra medium cuneatim angustatis, imo rotunde truncatis, marginibus cartilagineis subserratis cum punctis nigris in sinibus, flaccide chartaceis, supra pallide viridibus, nervis divaricatis fere recte adscendentibus paullo prominulis, venis tenuiter reticulatis, subtus subconcoloribus, nervis prominentibus, costa tenui striolata ; petiolo late planato, brevissimo : racemis elongatis, pendulis, cinereo-tomentosis; rachi subangulata; floribus sparsis pedi- cellatis; calycis limbo primum globoso, clauso, apice mucronulato, cinereo-tomen- toso, demum in lobos 3 vel 4 convexe rotundos longitudinaliter nervosos rupto; petalis 4, oblongis, utrinque attenuatis, glabris, albis; staminibus petalis longioribus ; disco styloque ut in char. gen.; ovario infero, turbinato, 4-loculari, ovulis plurimis ab axi superne funiculis suspensis: fructu elongato-oblongo, utrinque obtuso, ambitu rotundo (пес 4-gono), eduli. In insulis Oceani Pacifici: v. s. in hb. Mus. Brit. et Hook. Fiji, Vutu-kana (Seemann 150), Viti (Seemann); v. fr. in Museo Kew. (Seemann). A species near B. intermedia, differing in its broader leaves, with dissimilar nerves, its tomentose raceme with larger flowers. It is a tree 80-40 feet high, of erect growth; its leaves are 9-14 inches long, 34-5 in. broad, on petioles 1-2 lines long and broad; the raceme is about 10 in. long; rachis 1 line thick; flowers 1-1 in. apart ; pedicels 2 lines long; calycine limb in bud 4 lines in diameter; petals 11 in. long, 5 lines broad ; stamens 14 in. long: the fruit in Seemann’s specimen is 23 in. long, 14 in broad, crowned by the persistent calyx, and distinctly rounded, not acutely 4-gonous, as in the two other species growing in the island; the largest he saw, ina more mature state, was 4 in. long, 15 in. across, Тһе seed is eaten by the natives either when cooked or in the raw state, while in the other species it is poisonous. Seemann wrongly describes the calyx to consist of 4 rounded sepals, and did not notice that it is at first globose and entire, afterwards splitting into 3 or 4 lobes. It is called vutu-kana or vutu-kata. 13. Buronica cALYPTRATA, В. Br. MSS., in Benth. Austr. р. 287: ramulis crassis, cortice rimoso, e foliis lapsis cicatricibus majusculis creberrimis signatis: foliis superioribus mediocribus, reflexis, elongato-oblongis, apice ovatis aut rotundato- obtusis, imo sensim angustioribus, late acutis, marginibus tenuibus plicato-undulatis vix crenulatis, flaccide chartaceis, supra luride viridibus, in areolis glauco-opacis, nervis tenuibus rufulis prominulis reticulatis, subtus ferrugineo-pallidioribus, opacis, | _ costa striolata nervisque paullo prominentibus; petiolo compresso, limbo 24plo | | breviore: racemo vix terminali; rachi crassiuscula, angulata, striata; floribus | brevissime pedicellatis ; calycis limbo primum globoso et integre clauso, rugose | costatim nervoso, in lobos 2, dein in 4 rupto; petalis subparvis oblongis; stamini- | МБ. 7. MIERS ON THE BARRINGTONIACE Ж. riri bus numerosis, imo breviter monadelphis; disco styloque ut in char. gen. ; ovario infero, turbinato, 4-loculari, ovulis pluribus axi suspensis. In Australia: v. 8. in hb. Mus. Brit. Lizard Island, New South Wales (Banks апа Solander). A very distinct species. Branchlets 1-2 thick, singularly marked with crowded cicatrices of the fallen leaves; leaves 5-7 in. long, 2-22 in. broad, on petioles 21-33 lines long; rachis of spike more than 8 in. long; flowers about 2 in. apart; pedicels 4 line long; calyx in bud globose, 4 lines in diameter, lobes very thin and brittle, obtusely oblong, little longer than ovary; stamens very numerous, not very long, all anthe- riferous. 14. BuTONICA CEYLANICA, nob.: Barringtonia ceylanica, Gardn.: Barringtonia race- mosa, Thwaites (non Linn. nec Roxb.), Enum. 119 (excl. syn. et icon. Wight): Stravadium obtusangulum, Bl. in Van Houtte, Fl. Serr. vii. p. 24: ramis crassis, pallide brunneis, vel cinereis: foliis in ramulis brevibus congestis, oblongis, apice sensim acutis aut in acumen breve subito constrictis, imo cuneatis, crenato-serratis, dentibus заре aciculatis, curvatim subplicatis, flaecide chartaceis, supra fusco- viridibus, opacis, nervis tenuibus remotiusculis divaricatis еб valde arcuatim nexis, venisque transversis reticulatis prominulis, subtus flavide pallidioribus, nervis rufis venisque prominentibus ; petiolo subplano, fusco, limbo 24plo breviore: racemis ter- minalibus, mox pendulis, folio 2-3plo longioribus ; rachi tetragona, angulis rugosis, alternatim spicatiflora; pedicellis fuscis, tenuibus, calyce longioribus; calycis limbo ovali-globoso, integre clauso, demum in lobos 3 rotundatos rupto; petalis 4, triplo longioribus, oblongis, lateribus revolutis; staminibus, disco styloque ut in char. gen.; ovario infero, turbinato, tetragono, 4-loculari, ovulis in quoque loculo 2, suspensis: drupa immatura elongato-oblonga, utrinque subacuta, calyce coronata, abortu 1-loculari et l-sperma. Іп Indis penins. et Ceylon: v. в. in hb. Hook. et Mus. Brit. Ceylon (Thwaites 2682); in hb. Hook. Indis penins. (Rottler). This species, peculiar to Southern India and Ceylon, grows in moist situations at an altitude of 1500 feet. Its leaves аге 6-8 in. long, 23-3 in. broad, always acute at base, upon а broad stout petiole 3-4 lines long, and with about 15 pairs of nerves ; slender raceme 19-15 in. long; slight angular pedicels about 1— in. apart and 5-7 lines long ; calyx in bud 3-4 lines long ; petals 8-9 lines long, 5 lines broad, internal margin of disk more urceolate. The species is distinguishable from Butonica racemosa (to which it has been referred) by its smaller, more elliptic, less acuminated leaves, longer petioles, flowers smaller, on comparatively longer pedicels, an elongated fruit acute at both extremities : it is nearer Butonica intermedia. Mr. Thwaites makes two varieties of this species, principally because in the one the calyx is quite closed, and in the other slightly open at the apex. This appears to arise from the different ages of the raceme. In the younger buds the calyx is of course quite closed; in the more advanced stage the limb begins to Open in the apex, and the pedicel lengthens. The thickish limb of the calyx splits irre- Sularly nearly to the base into 3 or 4 lobes. Тһе fruit, in section, is seen in one of Mr. Thwaites’s specimens ; it appears quite smooth, obtusely oblong, 1$ in. long, + in. SECOND SERIES.—BOTANY, VOL. I. M 78 MR. J. MIERS ON THE BARRINGTONIACE.JE. broad, with a pericarp 1 line thick, very different in size and shape from that of В. racemosa. 15. BuronicA CAFFRA, nob.: Barringtonia caffra, E. Mey. MSS.: Barringtonia race- mosa, Oliv. (non Bl.) in Flor. Afr. ii. 438 : ramulis crassiusculis, striolatis, fistulosis : foliis late vel elliptico-oblongis, apice sensim aut breviter acuminatis, infra medium gradatim angustioribus subcuneatis, marginibus cartilagineis subrevolutis, aut integris et crenulatis, vel punctis callosis obsolete dentatis, chartaceis, supra lote viridibus, nervis tenuibus patentim divaricatis et arcuatim nexis, subtus palli- dioribus, costa valida, deorsum sensim incrassata, nervis venisque reticulatis pro- minentibus; petiolo semitereti, latiusculo, brevi: racemo terminali, longiusculo, pendulo; тасы subtenui angulata, striata; floribus spicatis, speciosis, subsparsis, pedicellis brevibus; calycis limbo primum globoso et integre clauso, demum in lobos seepius 4 rupto; petalis 4, duplo vel triplo longioribus, oblongo-ovatis; sta- minibus, disco styloque ut in char. gen.; ovario infero, turbinato, subtetragono, . ruguloso, 4-loculari, ovulis in quoque loculo 2 suspensis : fructu ovato-oblongo, lobis calycinis immutatis coronato, obsolete tetragono. In Africa australi: v. s. im Ab. Hook. fluv. Zambesi (Miller), fluv. Rovuma (Miller), fl. Luaba (Kirk), Zanguebar (Kirk 78), S. Africa (Drége) ; v. fruct. in Mus. Kew. (Kirk 78). A large tree, according to Miller, who found it growing along the banks of the Rovuma over an extent of twenty miles. It is very distinct from the Indian species В. racemosa. The branches are closely covered at the extremities by large leaves, which soon fall off and leave large cicatrices at the points of their insertion 3 lines apart; the leaves are 8-17 in. long, 23-43 in. broad, on petioles seldom more than 1 line long and broad, and have about 18 pairs of rather straight divaricating nerves; the raceme is 18-30 in. long, the pedicels about 2 in. apart and 2—8 lines long; the calycine limb is globular and entire in the bud, is 3—4 lines in diameter, the petals about 9 lines long. Тһе fruit col- lected by Dr. Kirk is oval, slightly narrowing at each extremity, obsoletely 4-angled, erowned by the unchanged calycine lobes, is 2 in. long, 11 in. broad, and contains а single seed, as in other species of this genus. The leaves in Drége’s specimen are longer and broader than in the others, and probably may be regarded as a variety. 16. BuTONICA APICULATA, nob.: ramulis crassiusculis, rugosis, striatis, summo стерге foliosis : foliis oblongis, apice in acumen breve acutum subito constrictis, imo cuneatis, obtuse et breviter serratis, chartaceis, supra viridibus, nervis tenuibus utrinque circiter 10 cum aliis brevioribus interjectis, prominulis, costa tenui, subtus paullo pallidioribus, subopacis, costa prominente ; petiolo limbo 15-20plo breviore : racemo terminali, brevi; rachi angulata, compressa, striata, opaca ; floribus approximatis, breviter pedicellatis ; calyce globoso, integro, demum in lobos 4 rupto ; ovario infero, 4-loculari, ovulis 2 in quoque loculo suspensis : fructu immaturo oblongo, 4-angulari, utrinque subaeuto, lobis calycinis 4 coronato. Іп Madagascar: v. s. in hb. Hook. loc. cit. ( Pervillé). A species collected by Capt. Pervillé in Madagascar. It is distinguished from the MR. Л. MIERS ОМ THE BARRINGTONIACE Ж. 79 preceding by its smaller leaves sharply spiculated, with fewer nerves and a longer petiole, a much shorter and stouter raceme. Its leaves are 5-6 in. long, 24-23 in. broad, on a petiole 8—5 lines long; the rachis of the raceme is 13 in. long, and bears alternate flowers 2-1 in. apart, on pedicels 2 lines long. The flowers had all fallen away, leaving some half-grown ovaries or immature fruits 7 lines long, crowned by 4 calycine lobes 3 lines long. 4. STRAYADIUM. This genus is maintained by very efficient and unmistakable characters. It was established by Jussieu іп 1709! upon the Zsjeria Saamstravadi of Rheede’ and a few other species. It was acknowledged by Blume? and DeCandolle *, but rejected by Roxburgh and other botanists, who conjoined it with a group of species under a genus called Barringtonia, very different from the true Barringtonia of Forster; and hence have arisen the many complications that followed. Miquel attempted a monograph: of the family?, which has created still further confusion among the species. He enumerated under Barringtonia 17 species*, apportioning them into 2 sections, Butonica and Stravadium, after the example of Endlicher', by characters so loose that he quite disregarded them in the arrangement of his species, and he thus rejected Stravadium as а genus. Blume at a later period stil acknowledged Stravadium*, enumerating 11 species, the first 6 of which only can be retained. The authors of the * Genera Plant- arum ' adopted entirely? the views of Miquel. We must attribute all these discrepancies to the entire disregard shown by most authors to the structure of the free portion of the calyx. Тһе difference is well marked in several cases, where this free portion is vesicular and entire in the bud, but is soon ruptured by the pressure of the petals into 2, 3, or 4 lobes, which are either very large or moderately small—characters constant in the genera Barringtonia, dgasta, and Butonica ; on the other hand, in Stravadium this calycine border is divided, even in the bud, into 4 small rounded ciliolated sepals, somewhat imbricated in sestivation, as in Carya and Planchonia. Tt is easy, therefore, without the possibility of mistake, to detect at once any species of Stravadium, a genus distinguished by the presence of 4 free sepals in the calycine border, an ovary constantly 2-celled, with 2 suspended ovules in each cell, producing an oblong, subtetragonous, indehiscent fruit, which by abortion is 1-celled, with à single large seed that fills the cell, and that consists of an exalbuminous embryo, like that already described іп Butonica. se : We may regard the Eugenia acutangula of Linnzus as the type of the genus, with Which many other different species have been confounded by botanists under the name of Barringtonia acutangula. ' Gen. р. 326, * Hort. Malab. iv. р. 15, tab. 7. * Bydr. 1097. deg * Prodr. 3239. 5 Fl. Ned. Ind. i. 485. * Van ан Fl. Serr. vii. р. 23. * Gen. РІ. p. 1233. no. 6325. * Van Houtte, Fl. Serr. vii. p. 23. э Gen. Pl. i. 720. M2 80 МВ. 7. MIERS ON THE BARRINGTONIACE Ж. STRAVADIUM, Jussieu. Stravadia, Pers. ; Meteorus, Lour.; Butonica (in parte), Rumph. ; Barringtonia (in parte), Miquel et alior.; Botryoropis, Presl. Calycis adnati limbus 4-sepalus: sepala parva, rotundata, membranacea, sæpius ciliata, æstivatione sub- imbricata, persistentia. Petala 4, oblonga, 3—4plo longiora, unguibus tubo staminigero affixa et cum illo caduca. Stamina numerosissima, multiseriata, imo breviter in tubum erectum monadelpha: filamenta filiformia, petala vix excedentia, qualia, omnia antherifera. Stylus tenuiter subulatus, paullo incurvus, eorum longitudine : stigma parvum, clavatum : ovarium inferum, turbinatum, szpius 4-angulatum, semper 2-loculare, vertice intra discum valde concavum ; ovu/a in quoque loculo 2, ab apice funiculo brevi suspensa. Fructus oblongus, plus minusve elongatus, tetragonus, angulis ssepius costatis, calyce coronatus, abortione 1-locularis et monospermus : pericarpium subcrassum, coriaceum, fibrosum : semen loculum implens, apice suspensum, homogeneum, structura ei Butonice simile. Arbores in India orientali, Malacca, Archipelago Asiat. et Australia vigentes, sepius mediocres : folia ob- longa, utrinque acuta, serrata, petiolata : racemi terminales, longe spicati : flores subparvi. 1. STRAVADIUM ACUTANGULUM, nob.: Eugenia acutangula, Linn. Sp. Pl. i. 673 (excl. syn.); Syst. Veg. i. 461; Flor. Zeyl. p. 85. no. 190 (excl. synon. in Pl. Her- mann. iv. p. 50); Willd. Sp. ii. 966 (excl. syn.): Barringtonia acutangula, Gaertn. Fruct. ii. 97, tab. 101; Roxb. Fl. Ind. ii. 635 (exclus. syn. Lour. et Rumph.); Thwaites, Enum. p. 118 (in parte): foliis in apice ramulorum congestis, oblongis, apice obtusis aut subacutis, infra medium sensim cuneatis, chartaceis, margine crenato-serrulatis, dentibus obtusis ssepe aciculato-mucronatis, glabris, supra viri- dibus, minute granulosis, nervis tenuissimis venisque reticulatis semiimmersis, subtus cinereo-glaucis, opacis, nervis venisque prominulis; petiolo semitereti, supra plano, subtus corruguloso, limbo 10plo breviore: racemo terminali, spicato, mox pendulo, folio 6plo longiore; floribus parvis, subsparsis, brevissime pedicellatis, glabris; sepalis 4 (rarissime 3), ovatis, erectis, marginibus membranaceis acute serrulatis; petalis 4, obovatis, 3plo longioribus, membranaceis; ovario infero, turbinato, acute 4-angulari, 2-loculari: fructu (sec. Gaertn.) oblongo-ellipsoideo, tetragono, angulis crasse costatis, utrinque attenuato, calyce coronato, 1-loculari et monospermo, semine ut in char. gen. et 4-partibili. In Ceylonia: v.s. in hb. Mus. Brit.; in hb. Hermanni, loc. cit. specimen typicum Linnei, Ceylon (Thwaites 1572). In the typical specimen the leaves are 11-21 in. long, 2-11 in. broad, on petioles 2-3 lines long; raceme 13 in. long ; pedicels 1 line long; sepals 2 line, petals 24 lines long. Gaertner’s figure of the fruit is from a specimen in the Leyden Museum, which came from Ceylon under the vernacular name Medella, the same as mentioned by Thwaites, and was probably contributed by Hermann to the Leyden collection, as he published there his work on the Ceylon flora. It is 13 in. long, 2 in. broad across the prominent angles ; the seed is oblong, subtetragonous, rounded at base, subacute at the apex, 13 lines long, 6 lines broad, the external portion (exorhiza) splitting into 4 parts along the sides. The plant in Ceylon bears the vernacular name of Ella-medella-gass. р M of the typieal plant, its flower, fruit, and seed, are shown in Plate XVII. MR. 7. MIERS ON THE BARRINGTONIACE Ж. 81 2. STRAVADIUM OBTUSANGULUM, Blume, іп V. Houtte, Fl. Serr. vii. p. 24 (excl. syn.) : Barringtonia obtusangula, Miq. Flor. Ned. Ind. i. 491: Barringtonia acutangula, Thwaites (non Gaertn.) (in parte), Enum. p. 119: ramulis brunneis, remote lenti- cellatis: foliis oblongo-elliptieis, apice obtusis, raro vix acutioribus, imo sensim cuneatis, retuse serratis, dentibus obtusis, subrigide chartaceis, supra fusco-viridibus, costa tenui, nervis subadscendentibus venisque granulatis valde reticulatis prominen- tibus, subtus ferrugineo-opacis, costa nervisque prominentibus ; petiolo fusco, sulcato, late marginato, limbo 12plo breviore: racemo terminali, spicato, folio longiore; rachi tenui, subflexuosa, angulata, glauco-pruinosa; floribus subparvis, subapproxi- matis, pedicellatis ; sepalis 4, rotundatis, carnosis, margine membranaceo minute dentieulatis; petalis 4, oblongis; ovario infero, turbinato, acute 4-angulato, rugu- 1050, fusco: fructu (sec. ВІ.) elongato-ellipsoideo, utrinque subattenuato, obtuse tetragono, angulis incrassatis. In Ceylonia: v. s. in hb. Mus. Brit. Thwaites, 1592: Sruct. non vidi. This species differs from the preceding in its longer, broader, more obversely elliptic, much darker, more glabrous leaves, of very different aspect. "They are 3-5 in. long, 15-2 in. broad, on petioles 3—4 lines long; its raceme is 3-10 in. long; flowers 2-3 lines apart, sometimes more aggregated, upon pedicels 1-2 lines long; sepals $ line long; petals 2 lines long. Thwaites’s specimen agrees well with Blume’s description, drawn from a Ceylon plant collected by Royer. Miquel states in a haphazard way that it is iden- tical with the Planchonia valida of Blume, which, according to the former, is merely a form of Planchonia sundiaca. This cannot be admitted; but it offers another proof of the continual contradictions between Miquel and Blume which serve to mislead us. 3. STRAVADIUM DEMISSUM, nob.: arbor demissa; ramulis tenuibus, pallide brunneis, » striatis, apice confertim foliiferis: foliis elliptico-oblongis, seepe subplicatim recur- vulis, apice obtusatim acutis, imo sensim cuneatis, margine cartilagineo vix reflexo leviter sinuato-serratis, dentibus obtusis intus mucronulatis, firme chartaceis, supra _ opace viridibus, sub lente minutissime granulosis, nervis remotiusculis subadscen- dentibus semiimmersis reticulatis, costa tenui, subtus pallidis, opacis, sepe de- albatis, costa, nervis venisque albescentibus et prominulis; petiolo semitereti, mar- ginato, imo sæpe crassiore, limbo 12plo breviore : racemo terminali, sæpe longis- simo, pendulo; rachi tenuissima, fusca, striolata, spicatiflora ; floribus pro genere majoribus, sparsis, pedicellis bractea longiore refracta cuneato-lanceolata acute denticulata membranacea glabra munitis ; sepalis 4, rotundato-ovatis, fuscis, sparse puberulis, marginibus membranaceis, ciliolatis; petalis 4, quintuplo longioribus, oblongo-obovatis, membranaceis, carneo vel pallide rubellis, extus subpruinosis. unguibus tubo staminigero extus adglutinatis ; ovario infero, turbinato, acute tetra- gono, 2-loculari, ovulis in quoque loculo 2, apice suspensis : fructu immaturo oblongo, imo obtuso, apice calyce coronato, acute quadrato, angulis subcrenatis, abortu 1-loculari et monospermo. In Malacca et in insulis Malayensibus: v. s. in herb. Hook. Assam (Masters 363), Tenasserim et Andaman (Helfors 2422-2425) Ma- lacca (Griffith), Malay penins. (Griffith 74), hort. cult. Cale. (Griffith 2425), 82 MR. J. MIERS ON THE ВАККІМСТОМІАСЕ Ж. Mergui (Griffith); in herb. Soc. Linn. Amherst (Wall. Cat. 3635 с); Guayalpoor (Wall. Cat. 3635 А), Chittagong (Wall. Cat. 3635 т). This species is said by Griffith to be a low tree, growing in moist woods, over all the Malayan provinces. Its branches are slender, with leaves crowded at their extremities, with axils 2-3 lines apart. The leaves аге 33—5 in. long, 13—2 in. broad, on petioles 3-5 lines long; the raceme is 12-18 in. long, on a rachis 1-3 line thick, with flowers 4-5 lines apart, on slender pedicels 14 line long, supported by a linear persistent bract 2 lines long; the sepals are 1 line long and broad, petals 5 lines long. The immature fruit, in Griffith’s specimen, is 11 in. long, 8 lines broad, has 4 acute almost winged angles. The species offers much resemblance to S. Rheedii, but differs in being of low stature, producing longer spicated racemes, which bear larger flowers on pedicels bracteated at their base, and in its fruit with almost winged crenulated angles. 4. SrTRAvADIUM RHEEDII, Blume in Van Houtte, Flor. Serr. vii. p. 24 (excl. synon.): Barringtonia acutangula, Roxb. (non Gaertn.) in parte, Fl. Ind. iii. p. 635 (excl. syn.); W. & A. Prodr. Fl. Ind. p. 335 (excl. syn.) : Eugenia racemosa, ВохЪ. (non Linn.), Icon. E. I. C. Mus. tab. 149: arbor valida, ramosa, ramulis tenuibus, valde rugosis, pallide brunneis, striatis, cortice crasso, fusco, asperato : foliis subconfertis, elliptieo- vel spathulato-oblongis, apice obtuse acuminatis, raro acutioribus, imo euneatis, serrulatis, dentibus obtusis ssepe mucronulatis, rigide chartaceis vel sub- coriaceis, supra saturate viridibus vel pallidioribus, opacis, sub lente (preesertim in nervis) granulis minutis asperatis, nervis tenuibus paullo prominulis, eosta sub- immersa, subtus pallidioribus, opacis, seepe dealbatis, glaucis vel fulvidis, nervis costaque prominulis reticulatis; petiolo supra plano aut sulcato, marginato, subtus corrugulato, limbo 15-20plo breviore: racemo terminali, spicatifloro, pendulo, folio duplo aut ultra longiore ; rachi subtenui ; floribus subapproximatis, longiuscule a cellatis; sepalis 4, ovatis, membranaceis, ciliatis; petalis бро longioribus, 4, oblongis, carneis vel intense rubris, extus carinatis, marginibus ciliatis; staminibus coccineis; ovario infero, turbinato, acute 4-angulato, 2-loculari, ovulis in loculis suspensis: fructu oblongo, utrinque subangustiore, calyce coronato; pericarpio coriaceo, interrupte plurisuleato, acute tetragono, angulis tenuiter nerviformibus, abortu 1-loculari et monospermo ; semine oblongo, profunde parallele sulcato, solido. Arbores in Indis peninsula undique sparse: v. s. in herb. Hook. penins. Ind. (Wight 1083), Travancore (Wight 1062), Madras (Wall. Cat. 3635 в), N. W. Ind. (Royle), Serampore (Edgeworth), Afghanistan (Ritchie 286), in planit. Ganges (Thompson), in hort. Cale. cult. (Griffith 2494) ; in hb. Soc. Linn. hort. Cale. cult. (Wall, Cat. 3635 к), ex hb. Roxburgh (in Wall. Cat. 3635 F), ex hb. Finlayson (Wall. Cat. 3635 1) ; in hb. Sir J. Smith. The several plants from all parts of the Indian peninsula above indicated are here con- ой to be the same as the Barringtonia acutangula, Roxb.; but they may require Dhaene and кенен Locket upon this species as the Туегіс Saamstravadi of Rheede (Hort. Mal. tab. 7); but they were misled into this opinion by the circum- MR. J. MIERS ON THE BARRINGTONIACE. 88 stance, not noticed by them, that the Saamstravadi in that place is reduced to half its natural size: that plant is the Butonica rubra, nob. (ante, p. 70). At the same time, Roxburgh, Wight, and others considered this species to һе identical with the дета acutangula of Linnszeus and of Gaertner; but there is ample evidence to prove the contrary. This plant closely approaches 5. demisswm, a species apparently confined to the Malay provinces and adjoining islands ; but it differs in its smaller and paler flowers. Roxburgh says it is a large handsome tree, in appearance like a regular and well-shaped middle-sized Oak, with slender branches; its leaves vary in size from 3-6} in. long, 13-25 in. broad, on petioles 3-6 lines long; the raceme is 6-10 in. long, the flowers 3 lines apart, on pedicels 1-2 lines long; the sepals are 4 line long and broad, the petals З lines long; the immature fruit is 14 in. long, 1 in. broad; the embryo of the seed when in a fresh state has a savoury taste, but afterwards becomes bitter. 9. STRAVADIUM PUBESCENS, nob.: Barringtonia acutangula, W. & А. (поп Roxb. nec Gaertn.), Prodr. Fl. Ind. p. 335 (in parte et excl. synon.): ramulis subangulatis, rugoso-striatis, ex cinereo уе] rufescente brunneis, subpuberulis vel glauco-pruinosis : folis elliptico-oblongis, apice latioribus, obtusis aut brevissime apiculatis, imo sensim subacutis, заре cuneatis, subserrulatis, dentibus mucronulatis, chartaceis, supra pallide viridibus, nervis tenuissimis subimmersis, venis transverse reticulatis, subtus ssepe pallidissimis, opacis, costa nervisque prominentibus, brevissime tomen- tellis, preesertim in nervis; petiolo subcrasso, imo latiusculo, puberulo, supra plano, limbo 10plo breviore: racemis sepe іп ramulis novellis brevibus terminalibus, folio 3-6plo longioribus, rachi tenui, angulata, griseo-puberula, spicatiflora; floribus pedicellatis; pedicellis imo braeteola parvula squamiformi æquilonga munitis ; sepalis 4, parvis, rotundatis, extus puberulis, membranaceis, ciliatis; petalis 4, oblongis, triplo longioribus, pallide roseis, raro intensioribus ; ovario infero, turbinato, tetragono, aspere puberulo, 2-loculari. In Indis peninsula: v. в. in hb. Hook. Coro- mandel (Konig, Roxb. in Wall. Cat. 3635 в), Concan (Dalzell), Carnatic (G. Thoinp- son), Indie penins. (Wight 1062-1080), Courtallam (Wight 1063; in Wall. Cat. 3635, ex hb. Wight). This species, which seems confined to the more southern provinces of the Peninsula, is very near the preceding, differing among other characters in its singular pubescence. Its leaves are 21-5 in. long, 14-24 in. broad, on petioles 3-6 lines long; the racemes are 12-18 in. long, pedicels 1 line long, sepals $ line long, petals 3 lines long; the fruit is unknown. 6. SrRAYADIUM сосотхкпм, DC. Prodr. iii, 289: Meteorus coccineus, Lour. Coch. Chin. ii. 449: arbor magna, ramis tortuosis, adscendentibus, ramulis crebre striatis : foliis sub- sparsis, elliptico-oblongis vel obovatis, apice obtusis aut vix acutis, imo sensim cuneatis, ad marginem subintegrum obsolete crenato-serrulatis, rigide chartaceis, supra subhepaticis, opacis, nervis rufulis costaque plana subprominulis valde reti- culatis, subtus flavide opacis, sub lente minutissime granulatis, nervis flavescentibus < Costaque prominentibus; petiolo supra plano, limbo 12-16plo breviore: racemis 84 MR. J. MIERS ON THE BARRINGTONIACE#. terminalibus, folio triplo longioribus, pendulis; rachi gracili, hepatica; floribus alternatim spicatis; pedicellis tenuibus, angulatis, fuscis, sepalis triplo longioribus; sepalis 4, parvis, ovatis, pallidis, marginibus membranaceis, ciliolatis, extus pul- verulento-puberulis; petalis 4, triplo longioribus, oblongis, membranaceis, coccineis vel roseis, cum staminibus intense purpureis, disco styloque ut in char. gen. ; ovario infero, turbinato, octagono angulis 4 acutioribus, vertice intra disci urceolum latum concavato, 2-loculari, ovulis in quoque loculo 2, suspensis: drupa late ob- longa, obtuse octagona, glabra, fusca, calyce coronata, abortu 1-loculari et 1-sperma : semine magno, duro, rotundato. In Cochin-China : v. s. in hb. Mus. Brit. plantam typicam (Loureiro); in hb. Hook. Siam (Schomb. 182—309). A tree of considerable size, growing in woods, very distinct from Butonica terrestris, with which it has been confounded. Its leaves are 4-5 in. long, 13-2 in. broad, on petioles 8-5 lines long, and with about 10 pairs of nerves. Тһе racemes are 10-18 in. long, with flowers 3—5 lines apart, on pedicels 3 lines long, having a small deciduous bract at their base; the sepals are 1 line long, the petals 3 lines long; the size of the fruit is not stated. Loureiro describes the corolla as monopetalous, not observing that the petals were agglutinated to the staminiferous tube. 7. STRAVADIUM LUZONENSE, nob.: Botryoropsis luzonensis, Presl, Epim. Bot. p. 220; Miq. Fl. Ned. Ind. i. 492; Walp. Ann. ii. 642: ramulis dichotomis, subtenuibus, cinereo-brunneis, striolatis, apice confertim foliiferis: foliis oblongo-ellipticis, apice in acumen breve acutum constrictis, imo subcuneatis et in petiolo brevi decur- rentibus, marginibus basin versus reflexis, undique serrulatis, dentibus erectis, sub- uncinatis, apice acute mucronulatis, flaccide chartaceis, supra rufescenti-viridulis, opacis, nervis arcuatim adscendentibus reticulatis, subtus pallidis, opacis, nervis paullo prominentibus ; petiolo brevissimo, semitereti, acute marginato, limbo 30plo longiore: racemo terminali, longiusculo ; rachi tenuissima, striolata, spicatiflora, rufescente, ebracteolata ; floribus plurimis, pedicellatis ; sepalis 4, rotundatis, margine late pallideque membranaceis, setulis rufulis ciliatis; petalis 4, oblongo-ovatis, 4plo longioribus, marginibus membranaceis, erosulis ; filamentis hzec excedentibus, crenu- lato-crispatis; disco angustissime annulari, margine interno in tubum crassum erectum ore lato integrum expanso; ovario infero, turbinato, tetragono, angulis sepalis alternis, pedicelloque pilis rigidis rufulis puberulo, 2-loculari, ovulis 2 in quoque loculo apici dissepimenti affixis. In ins. Philippinis: v. 8. iz hb. Hook. et Mus. Brit. Laguna in Luzon (Cuming 653, 1268). I have given the above diagnosis from my own observations. The genus Botryoropsis of Presl, formed on Cuming’s plant (658), which I have carefully examined, differs in 10 respect from Stravadium. Не was certainly wrong in attributing to it opposite branches and leaves; the 2-celled ovary and the broad tubular expansion of the inner margin of the disk are present in the preceding and all other species; indeed the latter feature differs only in degree, and is therefore of no more than specific value; the stamens are very numerous, and at least 6-seried ; and the anthers differ in no respect from the others; the ovules are affixed to the summit of the dissepiment; in fine, there is no generic МБ. J. MIERS ON THE BARRINGTONIACE®. 85 distinction whatevever. This species differs from the preceding in little more than in its smaller and narrower leaves, which are 23-5 in. long, 1-2 in. broad, on petioles 1-9 lines long; the raceme is 9 in. long, the pedicels 1 line, the ovary } line, the sepals 1 line, and the petals 3—4 lines long. 8. STRAVADIUM sPICATUM, Blume, in DC. Prodr. iii. 289, et in Van Houtte, Flor. Serr. vii. 24: Barringtonia spicata, Blume, Bydr. 1097 ; Miq. Fl. Ned. Ind. i. 489; Vriese, Ned. Kruidk. Arch. iii. 41: ramulis rugosis, apice erebre foliosis: foliis longe ellipticis, apice sensim acutis aut breviter acuminatis, imo cuneatis, irregulariter crenato-serratis, dentibus obtusis aut subacutis, intus sepe acute mucronulatis, tenuiter chartaceis, supra viridibus, opacis, nervis tenuissimis divaricatis vix pro- minulis, venis reticulatis, subtus pallidioribus, opacis, costa striolata nervisque prominulis ; petiolo tenui, semitereti, limbo 20plo breviore: racemo terminali, lon- gissimo, pendulo, rachi tenui, angulata; floribus spicatis, suberebris vel spar- sioribus; pedicellis brevibus; sepalis parvis, ovatis, pallide membranaceis, den- ticulis rubellis acutis eiliolatis ; petalis 4, duplo longioribus, oblongis, unguiculatis, staminibus styloque ut in char. gen. ; ovario turbinato, tetragono, 2-loculari, ovulis in quoque loculo 2, pendulis: fructu (sec. cl. Blume) ellipsoideo, utrinque obtuso, tetragono, costato-rugoso. In Тау prov. Bantam, in sylvis paludosis: v. s. in hb. Hook. Java (Zollinger 534*), Java (De Vriese, sub Barringtonia spicata); fruct. non vidi. A species distinct from S. Horsfieldii and S. reticulatum. Its leaves are 3-5 in. long, 15-2 in. broad, on petioles 4-6 lines long. Their dimensions agree with those given by Blume, who adds that they are sometimes (the lower ones ?) 8 in. long and 3 in. broad; the slender rachis is 12—15 in. long, with rather small flowers, either sparsely or more thickly spicated. Blume asks whether this is the same as the Carya macrostachya of Jack, which certainly it is not, as the latter presents a thiek cylindrical rachis, with much larger flowers, and belongs to the genus Дохопипа. Miquel, with his usual care- lessness, first copies the diagnosis of Blume as to the dimensions of the leaves, but afterwards, in contradiction, gives a much smaller size to the leaves—2 in. long, 11-2 in. broad, and therefore almost orbicular in shape. 9. SrRAVADIUM HonsrIELDII, nob.: Barringtonia Horsfieldii, Miq. Fl. Ned. Ind. 1. 489: ramulis tenuibus, ligneis, subfistulosis, rugulosis, axillis approximatis : foliis obo- vatis, apice rotundatis vel obtusatis et in acumen obtusulum breve latum subito constrictis, imo acute cuneatis, inzequaliter serrulato-dentatis, dentibus sepe breviter spinuloso-mucronatis, subchartaceis, supra lete viridibus, opacis, nervis tenuissimis patule divaricatis paullo prominulis, subtus pallidioribus, nervis venisque reticulatis, prominulis; petiolo tenuissimo, semitereti, limbo 8plo breviore : racemo in ramulis brevibus alaribus terminali, spicato, folio 6plo longiore; rachi tenui, flexuosa, griseo- puberula, pluriflora; floribus alternis, sessilibus, imo bracteola decidua munitis ; Sepalis 4, distinctis, rotundatis, subito acutis, denticulatis, puberulis; petalis 4, . Ovato-oblongis, marginibus crenulatis 4plo longioribus ; staminibus pluriseriatis, SECOND SERIES.— BOTANY, VOL. I. N 56 MR. J. MIERS ON THE BARRINGTONIACE E. imo altiuscule monadelphis ; ovario infero, turbinato, subtetragono, biloculari, ovulis in quoque loculo 2, funieulis suspensis. Іп Java: v. s. im hb. Mus. Brit. et Hook. Prowatt (Horsfield). A species near S. spicatum, but with much smaller leaves. In the above specimens the leaves are 8-44 in. long, 11-21 in. broad, on petioles 5-6 lines long, with about 8 or 10 pairs of nerves ; rachis of raceme 12-18 in. long ; flowers frequently in pairs, 3—5 lines apart; sepals j line long; petals 4 lines long; ovary strigosely puberulous, yellowish. Miquel describes the leaves as being 23—4 in. long, j-1j in. broad, on petioles % in. long, with 7—10 pairs of nerves. 10. SrRAVADIUM GLOBOSUM, nob.: Gustavia globosa, Spanagh. Linn. ху. p. 204: Peri- gara globosa, Span. l. с. 204: ramulis subvalidis, fusco-opacis, striolatis: foliis oblongis, supra medium obtusis, apice in acumen breve subito constrietis, imo cuneatis, pallidissime opacis, crenato-serratis ; petiolo semitereti, supra plano, limbo 10plo breviore: racemis axillaribus, folio longioribus ; rachi gracili, pendula, pallide opaca, angulata, pilis albidis rigidulis scabridulis; floribus parvis, alternatim subses- silibus; sepalis rotundatis, membranaceis, dentibus rubidulis, fimbriatis, subpube- rulis: ovario turbinato, rigide puberulo, 2-loculari: fructu parvo, globoso, abortu uniloculari, monospermo ; semine globoso, ab apice suspenso. In Java: v. s. in hb. Hook. loc. cit. (Anderson), ex Hort. Bogor. cultam (sub nom. S. spicatum), cum fructu immaturo. Leaves much approximated at the ends of the branchlets, 6-61 in. long, 21—24 in. broad, on slender petioles 7-8 lines long ; acumen 2 lines long and broad; sepals 1 line long and broad; ovary 1 line long: fruit (probably immature) 8 lines in diameter; pericarp 1 line thick; immature seed 3 lines іп diam., shown on a longitudinal section in the specimen above cited. 11. STRAVADIUM GRACILE, nob.: Barringtonia acutangula, Benth. (non Gaertn.) Flor. Austr. iii. 288: arbor magna, ramulis tenuibus, apice creberrime foliosis: foliis elliptico-oblongis, apice sensim angustioribus, et in acumen breve obtusulum subito constrictis, imo gradatim cuneatis, margine breviter serrulatis, dentibus obtusis, callosis, flaecide chartaceis, utrinque pallidissimis, pruinoso-opacis, sub lente granu- loso-punctulatis, costa nervisque tenuissimis, subarcuatim divaricatis paullo pro- minulis, reticulatis, subtus costa striolata nervisque albescentibus prominulis ; petiolo tenui, planato, subtus striolato, limbo 36plo breviore: racemo terminali, spicato, longissimo, pendulo; rachi tenui, angulata, fusca, pluriflora; floribus pedicellatis; pedicellis longissimis tenuissimis e nodis globosis callosis albidis ortis, imo bractea majuseula euneato-oblonga foliosa, dimidio breviore, alteraque minima rotundata membranacea squamiformi munitis ; sepalis 4, ovatis, submembranaceis, denticulatis ; petalis 4, oblongis, concavis, submembranaceis, dentieulatis, rubellis; ovario infero, turbinato, fusco, acute tetragono, 2-loculari, ovulis in quoque loculo 2 suspensis: fructu ovato-oblongo, tetragono, angulis acutis utrinque sublobatis. In Australia septentrionaji: v. в. in hb. Hook. Victoria River (Mueller), Adam's range (Mueller). MR. J. MIERS ON THE ВАВЕТУСТОМ!АСЕЖ. 87 А large handsome tree: but there are probably two species here; for Mueller states that in one variety the larger perspicuous bracts are wanting, but the minute mem- branaceous rounded scales remain. Тһе leaves are 21—51 in. long, 1-2 in. broad, on petioles 3—4 lines long; the slender graceful raceme is 18 in. long, with flowers 2-3 lines apart; pedicels 6 lines long ; bracts 4 lines long, 2 lines broad; sepals 3 lines long ; petals 7 lines long; ovary 3 lines long; fruit 11 in. long, $ in. broad, pedicellated, of а brownish eolour when dried. 12. STRAVADIUM RETICULATUM, Bl. in Van Houtte, Flor. Serr. vii. p. 24: Barringtonia reticulata, Miq. Fl. Ind. i. 494: Barringtonia acutangula, Korth. (non Gaertn. nec Roxb.) in Kruidk. Arch. i. p. 296: ramulis pallide brunneis, opacis, striolatis : foliis oblongo-lanceolatis, apice sensim vel breviter acuminatis, imo cuneatim at- tenuatis, argute serratis, subcoriaceis, nervis utrinque circa 12 divaricato-adscen- dentibus, venis valde reticulatis, costa prominente, subtus pallidioribus ; petiolo sub- tenui, supra plano, limbo 10plo breviore: racemo terminali, longe spicato, pendulo, folio 4plo longiore ; floribus subapproximatis, pedicellis tenuibus, angulatis ; sepalis 4, ovatis, membranaceis, denticulatis, medio crassioribus, pube brevissima asperatis ; petalis 4, obovatis, 3plo longioribus, membranaceis, coccineis, integris, unguibus tubo staminigero adglutinatis, cæteris ut in char. gen.; ovario ovato, infero, rufulo, rigide puberulo, 2-loculari, ovulis in quoque loculo 2, dissepimento pendulis : fructu ellipsoideo, calyce coronato, acute 4-angulari, costis rugulosis. Іп Borneo in uliginosis et ins. Philippinis : v. s. in hb. Hook. Barmassing (Motley 53, 582, 120), ins. Phillipin. (Cuming 1835). It is described as a small tree, with long slender racemes adorned with scarlet flowers. The fruit, which I have not seen, is said to be edible. The leaves are 5—64 in. long, 13-2 in. broad, on petioles 6-9 lines long; the raceme is 18 in. long, with a slender subcom- pressed rachis, slender pedicels 3 lines long ; the sepals are 1 line long ; petals 3 lines long; stamens 4 lines long: the disk on its inner margin has a-sharp elevated edge. In Blume’s account there is a mistake in the length of the petioles, which are 3-3 in., as Miquel rightly states. 13. STRAVADIUM SERRATUM, nob.: Barringtonia serrata, Miq. ТІ. Ned. Ind. i. р. 488; Walp. Ann. iv. 851: foliis lanceolato-oblongis, apice in acumen breve obtuse acutum constrictis, imo cuneatis, marginibus imum versus repando-dentatis, summum versus subremote et argute spinuloso-serratis, flaccide chartaceis, nervis inferioribus tenui- oribus patulis, reliquis validis et erecto-patulis, venis reticulatis : racemis elongatis, subpendulis ; pedicellis sepalorum longitudine dupla; sepalis obtuse rotundatis et denticulatis ; petalis flavido-roseis; ovario turbinato, tetragono, angulis prominulis. In Java,prov. Tjimarra : non vidi. The leaves are 12-17 in. long, 31—44 in. broad, on petioles 13-2 lines long; length of raceme not given; pedicels 6 lines long; sepals 3 lines long; number of cells in ovary not stated. From its ciliated sepals, we may infer that it belongs to Stravadium. N 2 88 MR. J. MIERS ON THE ВАККІМСТОМІАСЕ Ж. 14. SrRAvADIUM REINWARDTII, nob.: Barringtonia Reinwardtii, Miq. in Fl. Ned. Ind. 1. р. 488; Walp. Ann. iv. 851: foliis obverse oblongis vel obovato-oblongis, apice acuminatis, imum versus sensim attenuatis, tenuiter coriaceis, undique subspinulose serratis, nervis costatis utrinque circ. 12 erecto-patulis, venis reticulatis; petiolo brevissimo : racemo longissimo, pendulo, rachi crassiuscula; floribus pedicellatis; sepalis 4, obtusis, ciliato-erosulis; petalis late ellipticis, obtusis, supérne tenerrime ciliolatis. In Java: «on vidi. Leaves 9—11 in. long, 3—4 in. broad, on petioles 12-2 lines long; raceme 21 feet long; pedicels 2-3 lines long; sepals about З lines long; пена nearly 1 inch long. 15. SrTRAVADIUM COSTATUM, ВІ. in Van Houtte, Fl. Serr. vii. p. 24: Barringtonia costata, Miq. Fl. Ned. Ind. i. p. 489: foliis lanceolato-oblongis, apice acutis vel breviter acuminatis aut obtusioribus, imo cuneato-attenuatis, serrulatis, nervis costulatis; petiolo limbo 12plo breviore: racemis longissimis, pendulis; floribus pedicellatis, pro genere parvulis; petalis rubris: fructu obconico, subtetragono, costato. Іп ins. Sundaicis, sylvis humidis depressis: non vidi. А species said to differ from $. spicatum in its smaller pedicellated flowers, and fruit attenuated at its base. Its leaves are 3-65 in. long, 1-21 in. broad, on petioles 3-6 lines long; racemes more than a foot long. Miquel has referred here the Careya macrostachys of Jack; but this mere guess cannot for a moment һе entertained. 16. STRAVADIUM LUCIDUM, nob.: Barringtonia nitida, Miq. in Fl. Ned. Ind. i. 490: foliis obverse oblongis, obtusis vel obtusiusculis, imo attenuatis, irregulariter sinuato- serrulatis, tenuiter chartaceis, supra nitidis, costa tenui, nervis adscendenter divari- catis utrinque 8-12 tenuibus paullo prominulis, subtus pallidis, opacis, costa striolata prominula compressa, nervis venisque transversim retieulatis prominulis; petiolo semitereti late marginato, limbo 20plo breviore: racemis plurimis, spicatis, longis- simis, pendulis, patule griseo-pubescentibus; floribus sessilibus; calycis limbo po- culiformi in sepala 4 obtuse ovata diviso, pubescente, marginibus glabris membra- naceis eroso-ciliatis; ovario adpresse hirtello. In Java: v. s. in hb. Mus. Brit. Bantam, prov. Sourabaja (Horsfield). Horsfield's specimen agrees in character with Miquel’s description of this plant (which is also from Bantam), only that the leaves are somewhat larger, and probably from а more matured specimen. They appear identical; but I have changed the specific name, that it may not be confounded with the Planchonia nitida, Bl. In Horsfield's plant the leaves are 33—74 in. long, 21—3$ in. broad, on petioles 3-5 lines long; the racemes have all fallen ой; only the bottom of one remains, bearing a single half-grown fruit. Miquel states that the leaves are 31—5 in. long, 11-12 in. broad, on petioles 2-3 lines long; the racemes are 12-18 in. long. . It SrRAVADIUM DENTICULATUM, nob.: Ня к pallidis, suleato-striatis; foliis lanceolato- vel longé oblongis, apice sensim acutis, imo cuneatis, margine suberebre MR. J. MIERS ON THE ВАВВІМСТОУХІАСЕ Ж. 89 serrulatis, chartaceis, supra pallide viridibus, opacis, nervis paullo divergentibus, subimmersis, tenuiter reticulatis, subtus pallidissimis, opacis, costa tenui, albida, prominula, nervis pallidissimis vix prominulis ; petiolo angusto, utrinque marginato, subtus striolato, limbo 16plo breviore: racemo terminali, folio longiore, pendulo, glaberrimo, rachi subtenui, spieatiflora; pedicellis tenuissimis, 4-angulatis, sepalis longioribus; sepalis 4, ovatis, submembranaceis, marginibus undulatis, ciliato- denticulatis; petalis 4, triplo longioribus; staminibus numerosis, petalis multo longioribus, 4-seriatis; disco extus tubum staminigerum fulciente, interne in urceolam erectam ore latam et undulatam producto; stylo subulato, tenui ; stigmate parvo, bilobo; ovario infero, parvo, acute 4-angulari, 2-loculari. In Australia: v. s. in hb. Hook. Cape York ( Hann 195). The leaves are 4-82 in. long, 13-3 in. broad, on petioles 3-6 lines long; raceme 11 in. long, with slender pedicels 4 lines long; sepals 1 line long; petals 3 lines long; stamens 7 lines long; style 7 lines; ovary 1 tthe long. 18. STRAVADIUM SEMISUTUM, nob.: ramulis teneris, striolatis: foliis in summo ramulo- rum subconfertis, oblongo-ellipticis, apice sensim acutis aut breviter constrictis, infra medium subcuneatis, subserrulatis; petiolo subtenui, limbo 20plo breviore: racemo terminali, longissimo, pendulo, rachi tenui, angulato-striata, spicatiflora ; floribus subsparsis, aut sepe per paria approximatis, pedicellis tenuibus, patulis, sepalis zequilongis ; sepalis obtuse oblongis, glabris, marginibus subrevolutis ciliatis, ‚ quorum 2 semper in unum latius semicoalitis ; ovario infero, tetragono, sub-8-costato, disco plane annulari, intus in urceolam erectam crenulatam producto, vertice hinc profunde cavato, stylo subulato glabro sepalisque coronato, 2-loculari, ovulis in quoque loculo 2 suspensis. In Insulis Navigatorum: v. s. in hb. Hook. Falaga, Samoa (Powell 46). : A species remarkable for the semiagglutination of two of its four sepals, which are rather larger and thicker than usual in the genus ; this seems to be a constant character, and apparently occurs in the following species from the neighbouring island of New Caledonia. The leaves, approximated on the branchlets, are 43-7} in. long, 2-22 in. broad, on petioles 3—4 lines long: the slender terminal rachis is gradually bent till it becomes pendulous, is 20 in. long; the flowers are 1-5 lines apart, on spreading pedicels 2 lines long; the sepals are 3 lines long, the style 24 lines long; the petals and stamens are all fallen off. 19. Srravapium INTEGRIFOLIUM, Montrouzier in Mém. Acad. Lyon, vol. viii. (1858) р. 309: Barringtonia Montrouzieri, Vieill. in Bull. бос. Linn. Normand. x. (1866) р. 11: arbor, foliis apice ramulorum confertis, obovato-oblongis, apice acumiuatis, imo cuneatis, integerrimis, marginibus undulatis, glabris, reticulato-nervosis, vix petiolatis : racemis axillaribus, longe pendulis, laxifloris, bracteis imo involucratis ; se- palis 2 v. 3 v. 4, subacute ovatis, concavis, marginibus apice coloratis ; petalis 4, sepalis longioribus, albidis; staminibus disco styloque ut in char. gen. ; ovario infero, turbi- nato, 2-loculari, ovulis in quoque loculo 2: fructu oblongo, compresso, szepe arcuato, 90 MR. J. MIERS ON THE BARRINGTONIACE#. 4-angulari, angulis sinuato-alatis, abortu 1-loculari et monospermo. In Nova Cale- donia et ins. Art, juxta rivulos: non vidi. A tree 26-33 feet high, growing in valleys; its leaves are 7-8 in. long, 23—8 in. broad, and subsessile; its racemes are 11-2 feet long, with laxly spicated flowers ; the fruit, with 4 very undulated wings, and crowned by the calycine lobes, is 5 in. long and 17 in. broad; the pericarp is fibrous, and, when the fruit has fallen to the ground, rots after a while, and а root protrudes like that of a radish. This species presents the peculiarity of the pre- ceding species : of its sepals, often entire, two or three are sometimes partly conjoined. 5. PGLANCHONIA. This genus, proposed by Blume, was first published in 1852, in Van Houtte's * Flore des Serres; he had, however, as far back as 1828, described its seminal features in De Candolle’s * Prodromus,’ upon a plant referred by him to Gustavia, the seminal characters of which were then unknown. Long prior to this, the first known species of this genus was found in 1770, in Queensland, by Dr. Solander, and afterwards, in the Gulf of Carpentaria, by the late Mr. Robert Brown, who in 1808 discovered it to be a new genus, and then described the plant under the name of Butonicoides crenata, in a memoir never published, but which is still preserved in the British Museum, He changed its name to Careya crenata in 1819, when Roxburgh first established the latter genus, of which the internal structure of the seeds was unknown. Planchonia agrees with Careya in habit and the form of its fruit, which also contains many seeds, enveloped in pulp; but the embryo in the former is dicotyledonous, as in Barringtonia, while iu Careya it is analogous to that of all other genera of the family. Brown's description of this genus is far more accurate and copious than that of Blume. Succeeding botanists gave very laconic characters of four or five species; but Miquel, in 1855, created con- fusion by combining them all into two species, with characters almost useless for the purpose of distinction, I have carefully examined Brown’s typical plant, and have compared this with his original notes. With the analysis I have made of other, more recent specimens from another part of Australia, and from those materials, the fol- _ lowing generic character is formed :— PLANCHONIA, Blume (olim Gustavia, ВІ.). Butonicoides, В. Вт. MS.; Careya (in parte), В. Вт. et Benth.; Eugenia, Soland. MSS. Sepala 4, distincta, rotundata, subcoriacea, æstivatione paullo imbricata, persistentia. Petala 4, triplo | longiora, cuneato-oblonga, reflexa, unguibus tubo staminifero agglutinata et cum Шо caduca. Sta- - mina numerosissima, pluriseriata, in tubum brevem monadelpha; filamenta filiformia, petalis lon- giora, wstivatione contortuplicata, nonnulla апаш ега. Discus epigynus, plane annularis, margine externo tubum staminiferum fulciens, interno in urceolum breviter erectum ore lato expansum; stylus tenuiter subulatus, stamina subequans. Stigma parvum, 4-lobum aut fimbriatum. Ova- | rium inferum, turbinatum, vertice intra discum concavum, 4-loculare ; ovula in quoque loculo plu- B rima, funiculis ad axem radiatim affixa. гира ovato-oblonga, utrinque obtusa, calyce coronata, extus leviter opace granulosa, abortu 3- 2- vel l-locularis: pericarpium molliter coriaceum, sub- _ Crassum, lignose fibrosum ; semina in loculis pauca vel plura, in pulpa pannosa nidulantia, funi- | MR. 7. MIERS ON THE BARRINGTONIACE Ж. 91 culo brevi lato munita, ovata, compressa ; infegumentum externum tenuiter membranaceum (В. Br.), evanidum ; albumen? (aut integumentum internum) album, crasse carnosum, subfirmum ; embryo inclusus ; radicula longiuscula, valida, teres, supra medium subito inflexa, et hine inzqualiter bi- crura, crure longiore raphz proximo ad hilum spectante, crure altero parallelo breviore cotyledones 2 gestante ; cotyledones brevissime, ovate, егесіге, applicitze, incumbentes, tenuiter foliaceze, marginibus integris (vel paucidenticulatis, R. Br., aut plicate ,Blume). Arbores in insulis Moluccensibus vel in Australia boreali vigentes, frondose ; folia alterna, oblonga vel orbiculata, crenato-serrata, glabra, petiolata; infloresceutia terminalis aut т ramulis novellis supe- rioribus, brevissime ramosa, hinc pseudo-paniculata ; flores speciosi. 1. PLANCHONIA CRENATA, nob.: Hugenia crenata, Soland. М8. : Careya crenata, R. Br. in MS. ined. no. 75, in Mus. Brit. (olim Butonicoides crenata, В. Br. l.c.) : Careya arborea, var. australis, Benth. (non Roxb.) in Flor. Austr. Ш. р. 288: arbor ramis expansis, iterum ramulosis, ramulis 3 ultimis floriferis, brevibus, crassiusculis, pal- lide brunneis: foliis late ovalibus, apice in acumen sspe lineare mucroniforme subito constrictis, imo in petiolum longiusculum breviter acutatis, margine crenu- lato-serrulatis, supra pallide viridibus, nervis tenuibus vix immersis, subtus paullo pallidioribus, costa carinata albicante, nervis prominulis ; petiolo tenui, acute margi- nato, limbo 4plo breviore: racemis subaxillaribus, solitariis, aut in ramulis novellis terminalibus, paucifloris; rachi valida, longiuscule nodosa; pedicellis brevissimis, ad nodos articulatis, 1-floris, flore hine sspe caduco; calycis limbo imo brevissime cupulato, cum sepalis 4 rotundate ovatis vel oblongis, coriaceis, fuscis, marginibus membranaceis subciliatis; petalis 4, oblongis, unguibus tubo staminigero aggluti- natis; staminibus numerosissimis imo in tubum brevem monadelphis; disco mar- gine externo staminifero, interno in tubulum brevissimum membranaceum undu- latum expanso ; stylo filiformi, valde incurvo ; stigmate parvo, obtuse lobato; ovario infero, turbinato, vertice intra discum valde concavo, 4-loculari, ovulis plurimis biserialiter axi affixis: drupa elliptica, oblonga, Ігеуі, viridi, sicca pallide brunnea, calyce styloque coronata; pericarpio crassiusculo, fibroso-coriaceo, 4-loculari, aut abortu loculo unico maturescente, polyspermo; seminibus compresse ovalibus, in pulpam pannosam immersis, integumento fusco membranaceo tectis ; albumine ? aut integumento interno? carnoso-coriaceo, primum viridi, demum albo; embryone ut in char. gen. In Australia: v. s. in Ab. Mus. Brit. Cape Grafton (Banks et So- lander), Carpentaria Bay (В. Brown) ; in herb. Hook. Port Darwin (Schomb. 847), Victoria River (Mueller): v. fruct. cum sp. R. Br. et Schomb. These ample details, derived principally from Brown's notes and specimens, are fully confirmed by all the others which I have seen. Brown describes it as a large or middle- sized tree, with a thick trunk and a spreading comose head; the branchlets are marked by numerous cicatrices of the fallen leaves, which are 1 in. apart; the leaves, thus some- what crowded, аге 4-44 in. long, 21-3 in. broad, on а slender petiole 9—10 lines long, and have about 16 pairs of parallel nerves, 3 lines apart. In Schomburgk’s specimen from Port Darwin, the inflorescence is axillary, short, with few flowers upon a rather slender rachis; the pedicels alternate, 2 lines long, articulated on a prominent node; the ovary, 92 MR. J. MIERS ON THE ВАККІМСТОМІАСЕ Ж. 3 lines long, is crowned by the tubular free portion of the calyx 1 line deep, surmounted by 4 erect, rounded, coriaceous sepals 3 lines long; the petals are 8 lines long, 3 lines broad. It is accompanied by a single loose fruit, 2 in. long, 11 lines broad, very ventri- cose on one side, straight on the other, owing to the abortion of 3 of its cells; the re- maining cell contains 4 seeds 4 lines long, with the hilum pointing to the straight edge; the pericarp, filled with woody fibres, is 3 lines thick. Brown’s specimen presents a very different aspect, owing to the thickening of the rachis, which supports a matured fruit and another only half matured, in which the petals and stamens still remain; the rachis is about 4 in. long, very stout, with long persistent nodes, which, as in Cowratari, sup- port each one flower articulated upon it; the pedicel, 1 line long, presents at its base a cuneated oblong bract 12 lines long, 4 lines broad. The half-grown fruit is oval, 12 lines long, 7 lines broad, supporting the calycine tube, 2 lines deep, with its 4 sepals, the petals and stamens still remaining. The ripe fruit is quite oval, 13 in. long, 12 in. broad, crowned by the free portion of the calyx and style; the contents of the 4-celled pericarp are nearly all destroyed by insects; the axis, decayed pulp, and many seeds, mostly damaged, are still manifest, confirming the description of Brown. А section of Schomburgk’s 1-celled fruit is shown in Plate IX. fig. 8; it is less destroyed by insects, and contains 4 horizontal seeds of the same shape and size, embedded in pulp, now reduced to loose grains; the outer thin integument has disappeared; but the inner thick coating (now white) remains; and in all four, at the apex, along one edge, there is a linear open slit, which extends to the cavity over the summit of the enclosed embryo; the embryo fills the cavity, is of the horseshoe-shape described by Brown; and I have given in fig. 14a magnified view of the same. Brown, in his description, says, “ albumen nullum;” but he described the thick inner coating (which, it appears to me, must be albumen) as being carnoso-coriaceous green when fresh, afterwards becoming white : it seems too thick for a _ mere integument, shows no sign of a raphe or chalaza; and being of a greenish colour, it responds more to the condition of albumen. Solander’s specimen has no inflorescence, u t has two loose flowers, which correspond with the others above referred to. 2. PLANCHONIA SUNDAICA, Miq. in Flor. Ned. Ind. p. 493 (excl. synom.); Walp. Ann. iv. 882: Planchonia undulata, Tenn. & Ben. in Cat. Hort. Bogor. : foliis elliptico- oblongis vel obovatis, apice acute acuminatis, imo cuneatis, marginibus cartilagineis, valde undulatis, subrevolutis, integris vel obsolete serrato-crenatis, rigide chartaceis; supra fusco-viridibus, sublucentibus, subrufescentibus, nervis plurimis tenuibus ru- fulis prominulis, costa tenui subimmersa, subtus pallidioribus, brunneis, costa rubra ‚ пегу1заще prominentibus, transversim reticulatis ; petiolo valido, marginibus latius- culis recurvatis, limbo 16ріо breviore: racemo terminali, brevi; rachi crassa, angu- lata, crebre multiflora; nodis floriferis apice bracteolatis; floribus ad nodos sub- sessilibus; sepalis 4, parvis, acute ovatis, subcoriaceis, griseo-pruinosis, imo in tubum brevem coalitis; petalis obovatis, 3plo longioribus, carnosulis ; ovario infero, cylindrico, 8-costato, vertice disco plano margine interno acute elevato, intus circa stylum valde concavo, 4-loculari, ovulis in quoque loculo pluribus, axi radiatim affixis : MR. J. MIERS ON THE BARRINGTONIACE. £. 95 drupa ellipsoidea, levi. In Java: v. pl. s. in hb. Hook. in hort. Bogor. cult. (De Vriese), eodem loco ( Benedyk, sub P. undulata) ; fruct. non vidi. The leaves have а reddish hue; they vary in size, 3-81 in. long, 13—4 in. broad, on a decurrent petiole 2-4 lines long; they have about 20 pairs of closely parallel nerves. The ovary is 2 lines long, the sepals 3 lines long, the petals 9 lines long, 4 lines broad, the many-seried stamens somewhat longer. 3. PLANCHONIA NITIDA, Bl. in Van Houtte, Fl. Serr. iv. p. 24 (exel. synon.) : foliis elliptico- oblongis vel ellipticis, apice acutis, acuminatis aut obtusis, imo cuneatis, margine insequaliter sinuato-serratis, chartaceis, supra nitidis; petiolo limbo 9-12plo breviore : inflorescentia ignota: drupa ellipsoidea, levi. Іп insulis Sundaicis: non vidi. This species must not be confounded with the Barringtonia nitida, Miq., which is a Stravadium. There is nothing in the above laconic description to show that it belongs to Planchonia. Its identity with Blume's original Gustavia valida cannot be admitted, as the more accurate Hasskarl has given so excellent a character of the latter plant. It must therefore be considered a very doubtful species of Planchonia. Blume says that its leaves are 3-84 in. long, 12—4 in. broad, on petioles 3-8 lines long. 4. PLANCHONIA TETRAPTERA, nob: Gustavia alata, Spanag. in Linn. xv. 204: Myrtus alata, Zeppell, 2. e. 204: Barringtonia acutangula, Spanag. 1. е. 204: Planchonia timoriensis, var. alata, Miq. in Fl. Ind. Ned. i. p. 493: JPlanchonia timorensis, Bl. in Van Houtte, Fl. Serr. vii. p. 25: ramulis junioribus alato-angulatis; foliis majus- culis lanceolato-oblongis vel oblongo-ellipticis, apice in acumen breve obtusum apiculatis, imo cuneatis et in petiolo late decurrentibus, margine irregulariter crenatis, dentibus lateraliter mucronulatis; petiolo alato, marginibus in ramulo decurrentibus, limbo 8-12plo longiore: inflorescentia ignota: drupa ovoidea vel ellipsoidea, Ігеуі. Іп ins. Timor: non vidi. It is better to change the specific name alata, that the species may not be confounded with Barringtonia alata, Wallich. Blume referred this species to his Planchonia timori- ensis, which he considered to be the same as the Gustavia alata of Spanaghoe, from Timor, with which the Myrtus alata of Zippel was regarded as identical: we recognize the same species in Miquel’s var. alata of Blume’s Planchonia timorensis. The species is distin- guished by the broadly winged angles of its branches. The leaves, according to Blume, are 32-8 in. long, 2-31 in. broad, on petioles 6-18 lines long, some of the leaves growing to а length ‘of 17 in. according to Miquel. Miquel’s variety bimensis may probably be a distinct species. ы 5. PLANCHONIA ELLIPTICA, nob.: ramulis brunneis, opacis, striolatis: foliis ellipticis, apice obtusis et in acumen breve subito constrictis, imo cuneatis, marginibus carti- lagineis inszequaliter crenulatis, rigide chartaceis, seepe conduplicatis, supra (in sicci- tate) fuscis, rubescenti-brunneis, lucidis, nervis parallele divaricatis tenuibus costaque tenui prominulis, subtus pallide brunneis, nervis venisque transverse reticulatis SECOND SERIES.— BOTANY, VOL. I. oe) 94 МВ. J. MIERS ON THE ВАККІМСТОМІАСЕ А. prominulis; petiolo fusco, revolutim marginato, limbo 8-9plo breviore: racemis in ramulis foliiferis, terminalibus, folio subbrevioribus; rachi erecta, crassiuscula, nodosa, nodis validis apice 1-floris (bracteatis ?); floribus sessilibus; calyce cupulato 4-lobo, lobis semiovatis erasse coriaceis, intus pruinosis, extus granosis, fuscis; petalis 4, sepalis 4plo longioribus, oblongis, parallele nervosis, pruinosis, pallide roseis, un- guibus tubo stamineo brevissime agglutinatis; staminibus numerosissimis, multi- seriatis, petalis subzequilongis, imo in tubum longiusculum erectum monadelphis, nonnullis antheriferis, reliquis sterilibus ; filamentis granulatim corrugulosis; stylo longo, tenui ; stigmate breviter fimbriato; ovario infero, turbinato, subtetragono, cor- . rugato, pruinoso, 4-loculari; disco epigyno plane annulari, margine exteriore tubum staminigerum gerente, intus fornicatim prominente, verticem concavum ambiente. In Borneo: v. s. in hb. Hook. Banjarmassing (Motley 750). The raceme-bearing branchlets 2 in. long, with 2—4 leaves erowded toward the apex; leaves 3—4 in. long, 1$-22 in. broad, on petioles 4—7 lines long, and with about 12 pairs of parallel nerves 3 lines apart; terminal raceme 2 in. long, furnished with many crowded very beautiful flowers, sessile on the nodose expansions of the rachis, which are 1 line long and thick, and 1-2 lines apart; ovary 2 lines long, its crowning sepals 3 lines long; petals 13 lines long, 43 lines broad; monadelphous tube of stamens 3 lines long, and 3 lines in diameter ; stamens numerous, many of them sterile ; ovary subtetragonous, 4-celled, with numerous ovules in each cell attached to the axis: their characters are quite those of Planchonia, to which this species is referred. Its fruit is unknown. 6. PLANCHONTA LrTTORALIS, Bl. in Van Houtte, Fl. Serr. уй. p. 25: Perigara valida, Bl. (in parte) Bijdr. Z. с. 1096 : ramis rugulosis, striatis, junioribus pallide brunneis, opacis, vix striolatis: foliis ellipticis, vel elliptico-oblongis, apice in acumen sub- breve subaeutum angustissime obtusulum et subemarginatum constrictis, imo sen sim et anguste cuneatis, margine subrevoluto cartilagineo inszequaliter sinuato serratis, dentibus erectis obtusis vel seepius acute mucronulatis, chartaceis, supra | pallide viridibus, opacis, costa tenuissima nervisque pallidis prominulis, venis valde reticulatis, subtus pallidioribus, costa striolata nervisque albescentibus paullo pro- minentibus ; petiolo tenui, limbo 15plo breviore: inflorescentia ignota : drupa elon- gato-ellipsoidea, costatim subangulata. Java in maritimis: v. s. in hb. Hook. in hort. Bogor. cult. (Anderson, sub G. valida, sine flore aut fructu). A species differing from P. sundaica in its broader, less acuminated, paler leaves, upon longer petioles. It is cultivated in the Bogor Gardens, probably introduced from one of the Sunda Islands. Its leaves, 2-1 in. apart, are 41-71 in. long, 11—4 in. broad, on petioles 3-6 lines long. Blume considered it the same as his Perigara valida, because it is so named in the Bogor Gardens; but that ifa very different species. 7. PLANCHONIA VALIDA, nob.: Pirigara valida, Bl. Bijd. 1096: Gustavia valida, DC. Prodr. iii. 290; Hassk. Bot. Zeit. (1844) xxvii. p. 595: procera: foliis ovali- vel late ellipticis, apice acute vel subretuse acuminatis, imo cuneatis vel retusis et in peti- ЫРДОО, ea МБ. J. MIERS ON THE BARRINGTONIACEJE. 95 olo decurrentibus, margine denticulatis; petiolo lato, limbo 12—14plo breviore, mar- ginato : inflorescentia ignota ; sepalis 4 semiorbicularibus, ovario zquilongis, erectis ; petalis 4plo longioribus, obtuse oblongis, viridiusculis, flaccidis, reflexis; staminibus quam hee longioribus, pluriseriatis, filiformibus, imo in tubum sepalis longiorem monadelphis, imo sanguineis, арісе albescentibus; disco epigyno plane annulari, extus tubum staminigerum gerente, margine interno in urceolum erectum expanso, vertice hine profunde cavo; stylo longe filiformi, flexuoso, viridi; stigmate obtuso ; ovario subturbinato, obsolete octagono, 3—4-loculari, ovulis in quoque loculo pluri- mis biseriatis. In Java et ins. Nusa Kambanga in sylvis : non vidi. The excellent description of Hasskarl shows that this species belongs to Planchonia, which is confirmed by the seminal structure indieated by Blume eighteen years previously. Blume, however, confounded it with his Planchonia nitide, from which it is evidently distinct. It is described as a tall tree growing in woods; its leaves are 3-7 in. long, 25-91 in. broad, on petioles 3-6 lines long, 3 lines broad ; the ovary is 3 lines long; the sepals of the same length; the petals are 14 lines long, 6 lines broad ; the staminiferous tube is 6 lines long, the filaments 18 lines long; the style is 24 lines long. 6. CAREYA. A genus established in 1819 by Roxburgh in his * Plants of Coromandel,’ and acknow- ledged by all botanists : it consists of four species, natives of India and the Malayan penin- sula, where they form trees of considerable size, with one exception, which is of low suffruticose stature. The inflorescence is extremely short, terminal on the nascent branchlets, with a thick fleshy rachis, bearing 1-4 very approximated sessile handsome flowers, each supported by 2 braets, often of large size: the adnate calyx is surmounted by а cup-shaped limb, divided halfway into 4 rounded segments, which are subimbricated in the bud ; 4 large obovate fleshy petals; the stamens, of unequal length, are very numerous, pluriserial and monadelphous at their base, and present the peculiarity that the outer and longer series, as well as the inner and shorter rows, are destitute of anthers, the intermediate series alone being antheriferous; the inferior ovary is usually 4-, rarely 5-celled, with several ovules in each cell, radiating from the central axis. "The fruit is globose, about the size of an orange, is crowned by the persistent calycine limb, has a smooth coriaceous pericarp, normally 4-celled ; but the dissepiments usually disappear, absorbed in the rather solid pulp, in which the many seeds are imbedded : the seeds are . oblong, subcompressed, about the size of a field-bean, sometimes smaller; they have no albumen, have an external thick testa, covering a solid mesopodal embryo, which often begins to germinate in the ripe fruit. This embryo, as in other genera, though solid, is actually formed of two agglutinated layers, the outer of which (the exorhiza) was declared by Roxburgh and Wight to be the albumen, the inner (neorhiza) being regarded by them ав ап embryo with agglutinated indistinguishable eotyledons and radicle. Вохһагыһ, іп à drawing copied by Wight, showed this seed in the act of germination; but this was afterwards better illustrated and explained by Dr. Thomson in the germinating seed of Careya arborea, where he showed there was no albumen, and that the inner body alone o 2 96 MR. J. MIERS ON THE BARRINGTONIACE А. (the neorhiza), by the protrusion of both extremities, produced at one end the growing leafy stem, at the other end the root of the new plant. Other observations of Dr. Thom- son are of extreme importance, as explaining the progress of this growth. No doubt can now be entertained on this subject. Griffith gave several figures to illustrate the growth of a similar seed in a plant named by him Careya pendula, but which I have shown to belong to the genus Рохотта. The Australian species considered a variety of Careya arborea is Planchonia crenata. САВЕХА, Roxb. Calycis adnati limbus breviter cupuliformis, fere ad basin in sepala 4 rotundata divisus, hzc in esti- vatione subimbricata, in fructu persistentia. Petala 4, calyce multo longiora, cuneato-oblonga, expansa, unguibus tubo staminigero affixis et cum illo caducis. Stamina numerosissima, pluri- seriata, petalis longiora, imo in tubum brevem erectum monadelpha : filamenta filiformia, insequi- longa, seriebus exterioribus longioribus, interioribusque brevioribus anantheris, intermediis solummodo fertilibus ; anthere parve, subbilobz. Stylus tenuiter subulatus, staminibus zequilon- gus incurvus. Stigma obsolete 4-lobum. Discus pulvinatim annularis, margine externo tubum staminigerum fulciens, interno in urceolum erectum expansus. Ovarium inferum, turbinatum, 4-loculare, vertice intra discum concavo; ovula in quoque loculo plurima, axi affixa. Fructus globosus, calyce persistente coronatus: pericarpiwm crasso-coriaceum, 4-loculare, dissepimentis mox evanidis; semina plurima, vaga, compresse ovata aut oblonga, in pulpam pannosam fibrillis interspersam immersa ; testa levis, pallide brunnea; chartacea, embryo exalbuminosus, elongato- oblongus, mesopodus, заре intra pericarpium germinans. Arbores frondose in India et Malacca indigene, rarius suffruticose : folia. sepe majuscula, oblonga vel orbiculata, serrata, petiolata : racemi terminales aut laterales subspicati ; flores speciosi, sessiles, imo sepius 2-bracteati. 1. САВЕХА HERBACEA, Roxb. Pl. Согот. iii. 13, tab.'217; Fl. Ind. ii. 688; DC.iii.295; W. & A. Prodr. Fl. Pen. i. 335; Wight, Icon. t. 557; Griffith, Proc. Linn. Soc. i. 280; Notul. pars iv. (excl. deser. pl. p. 660), tab. 684, cuj. figure 6 sistent (reliq. excl.) Miq. Flor. Ned. Ind. i. 494. In Ind. Or. : v. s. in hb. Mus. Brit. Nepal (Buchanan). A small half-herbaceous shrub, growing in the interior of Bengal and extending into Nepal. It has a ligneous perennial root, which throws out several short perishable branches. Its leaves are 4-8 in. long, 11-21 in. broad, cuneately obovate, acute at the apex, serrulate, smooth on both sides, on a petiole 4 lines long. It has short axillary racemes, bearing a few beautiful flowers, almost sessile, within 2 lanceolate bracts, which embrace the calyx; the calycine limb is 7 lines long, cup-shaped at the base, and divided halfway into 4 roundish lobes, which are distinct in the bud; the oblong pink petals аге 15 in. long, 5 in. broad; filaments closely many-seried, all agglutinated to the mona- . delphous tube, and falling off with the petals attached to its base, the exterior series bearing no anthers; ovary 4-celled, with 3 or 4 ovules in each cell. Fruit globular, less than an ineh in diameter, erowned by the persistent calyx, without any visible dissepiments, and filled with а greenish pulp, in which the seeds are enveloped; the seeds are oblong, compressed, 3 lines long, with a homogenous embryo, which often germinates within the fruit. | The description of Griffith, under the head of C. MEME bya mistake of the Editor, belongs to some Anacardiaceous plant. See my remarks under Doxomma pendulum (page 100). MR. J. MIERS ON THE BARRINGTONIACEA. 97 2. OAREYA ARBOREA, Roxb. Pl. Согот. iii. p. 14, tab. 218; Fl. Ind. iii. p. 638; Ham. Trans. Linn. Soc. xv. p. 97; Rheede, Hort. Malab. iii. p. 35, tab. 36; DC. Prodr. iii. 295; W. & A. Prodr. i. 334; Wight, Illustr. p. 20, tab. 99, 100; Miq. Fl. Ned. Ind. i. 494: Careya spherica, Wight, in parte (non Roxb.), Icon. tab. 556. In Ind. Orient.: v. pl. s. in herb. Mus. Brit. E. Ind. (Kónig), ibidem (Buchanan); in herb. Hook. Е. Ind. (Carey), Sikkim (Hook. et Th.), Afghanistan ad Kala Nudde (Ritchie 362). A tree, sometimes of immense size, inhabiting the valleys in Orissa and other provinces of India, with many spreading branches suleately angular, often covered with the cica- trices of fallen leaves. The broadly ovate leaves are suddenly constricted at the apex by a sharp point, are cuneate below the middle, the margins with small close teeth, sub- flaccid, pale green and opaque above, with about 16 pairs of slender prominulent nerves and a flat midrib, paler beneath, with straw-coloured prominent nerves; they are 7-8 in. long, 33-44 in. broad, on a channelled margined petiole 2-3 lines long. The raceme is terminal, 6-8 in. long, bearing about 8 flowers at the extremity of its stout rachis, which are almost sessile, with 3 bracts at their base; the calyx is tubular, 10 lines long, 5 lines broad, its margin being divided into 4 valvate acute lobes or sepals 3 lines long ; the petals are 12-15 lines long, 5 lines broad, with their margins laterally turned back ; the stamens are of various lengths, in many series, the longer and shorter series barren, the intermediate series antheriferous ; they are all agglutinated at their base into an erect tube, seated on the outer margin of the disk, and to which the claws of the petals are also affixed. The globular drupe is 2 in. in diam., with a smooth fleshy fibrous peri- carp 1 line thick; seeds ovate or oblong, compressed, scattered through a rather solid pulp. The fetid wood is useful, of a reddish colour, with a grain not so close nor so hard as mahogany ; and strong ropes are made from the fibres of its bark. З. CAREYA ярнжвтса, Roxb. Fl. Ind. iii. 336; Wight, Icon. (ex Roxb.), tab. 147 (non | 556); Мід. Fl. Ned. Ind. i. 494. In Malacca, Khasya, et Sikkim: v. s. in herb. Hook. Chittagong (Wall. Cat. 3640), Kumaon (Strachey), Sikkim (Hook. et Th.) Mergui (Griffith). A tree 30 feet high, with a trunk 12 feet long, 8 in. in diam., growing beyond the limits of the Indian peninsula. It yields a bark with strong fibres. It differs from C. arborea in its larger, broader, more rounded leaves, in its much longer racemes with a thicker rachis, in its more numerous, crowded, larger sessile flowers, bibracteolated at their base. Its branches have a smooth ashy bark, with leaves approximated on their summits. Leaves oblong-ovate, broad, rounded towards the apex, with a short abrupt acumen, suddenly and broadly cuneated towards the base, the margins being minutely crenulate-dentate ; they are subcoriaceous, lucid aboye, with about 20 pairs of patently divaricated nerves conjoined near the margin; they are 9-12 in. long, 43-67 in. broad, on a petiole 4-6 lines long. The racemes are terminal, several inches long, bearing near their summit 6-12 approximated sessile flowers, each supported by 3 basal rounded bractlets 3 lines long and broad; the ovary, 5 lines long, is surmounted by 4 nearly 98 МВ. J. MIERS ON THE BARRINGTONIACE.E. erect rounded sepals; the 4 petals subacutely oblong, 15 in. long, 4 lines broad, have the margins rolled back, and are of a palish green colour; ovary 4-celled, with numerous ovules in each cell. Drupe globular, nearly 2 in. in diam., crowned by the суете lobes; the pericarp and pulp, of a yellowish colour, contain few seeds com- pressed oblong, about 5 lines long, constructed like those of C. arborea. The specific name is not derived from the shape of its leaves, but from the spherical head of large flowers approximated at the extremity of the rachis. Miquel suggests that the species, so well described and figured by Roxburgh, should be expelled from the genus; but as he never saw the plant, and offers no reason for his suggestion, it cannot be admitted. Тһе analysis of this species is shown in Plate XVI. fig. 9 et seq. 4. CAREYA ORBICULATA, nob.: ramulis lævibus aut rugulosis, cortice crasso : foliis orbi- eulatis apice rotundatis et obsolete emarginatis, imo subito breviter euneatim con- strictis, ad marginem planum insqualiter grosse crenatis, dentibus brevissimis obtusis, ehartaceis, supra ex brunnescente viridibus, sublucentibus, nervis paucis divergentibus tenuibus paullo prominulis, subtus brunneis, opacis, nervis flavidis venisque reticulatis prominentibus ; petiolo plano, compresso, marginato, limbo 20plo breviore: гасето terminali, brevissimo, rachi angulato-suleata pauciflora; floribus approximatis, sessilibus; calycis limbo cupuliformi fere ad basin in sepala 4 rotun- data carnosa diviso, extus pruinoso vel rufo-tomentello; petalis 4, obovato-ob- longis, calycis бра longitudine, medio crassiusculis, marginibus late submem- branaceis, extus flav ide pruinosis; staminibus pluriseriatis imo in tubum disco in- situm monadelphis, filamentis granulato-rugulosis; disco annulari, plano, margine externo tubum staminiferum gerente, intus in prominentiam tenuem erectam ver- ticem ambientem expanso; ovario infero, tetragono, profunde transverse corrugulato et fere tubereuloso, 4-loculari, ovulis numerosis in axe undique collateraliter affixis: In regno Burmensi et penins. Malayensi: v. s. in hb. Hook. Mergui (Goes Ind., Or. (Carey); fruct. non vidi. The stout branchlets of this species are 1 in. thick, closely studded with small cica- trices of fallen leaves; the leaves are 5-51 in. long, 41—41 in. broad, on petioles 3-4 lines long: the fragment of the rachis of the terminal raceme is only 3 1 in. long, marked by the nodes where 4 flowers were affixed; the flowers are loose; the rugous 4-angular ovary 4—5 lines long; the superior limb of the calyx is divided into 4 rounded sepals 4 lines long and broad; the petals 15 lines long, 8 lines broad; the filaments, of equal length, are granulated and conjoined at their base into a tube, to which the petals are also attached by their claws. ‘This species is shown in Plate XVI. fig. 6. түн 7. Doxomma is a new genus ', distinguished from all the preceding by well-marked characters. Its species are lofty trees, with rather large, oblong, cuneated leaves, often upon unusually long petioles. The inflorescence is a very long spike of large handsome flowers, often much approximated and sessilé, sometimes bracteolated, upon a stout fleshy rachis ` So named from 20а (gloria), бина (adspectus), because of its splendid flowers. · маан Бора зав дааа: WO URP ЧИТ ee а ИРИ ЧИНУ даналы ЧЕРНИТЕ MR. J. MIERS ON THE BARRINGTONIACE. 09 which is erect or more frequently drooping from its great weight; the calycine limb is shortly cup-shaped, and divided nearly to its base into 4 rounded ciliated sepals, imbricated in the bud; the disk as in the other genera; the stamens are monadelphous, upon an erect tubular ring, which is seated on the outer margin of the disk; the fila- ments, long and slender, coiled in estivation, afterwards spreading in many series, аге all fertile, with small yellow 2-celled anthers; the petals are large and oblong; the style, as long as the stamens, is subulate, terete, with a minute clavate stigma; the inferior turbinate ovary, often 4-winged, is 4-locular, with 2 ovules in each cell, sus- pended from the summit by thickish funicles. The fruit is cylindrically oblong, furnished with 4 equal longitudinal undulating wings; the periearp is coriaceous and fibrous, by abortion 1-сеПей, with a single suspended seed, which fills its cavity; this is covered by a thin coriaceous testa, somewhat adhering to the pericarp; the enclosed embryo is white, nearly as hard as ivory, homogeneous, consisting of two layers, separable by a stratum of tissue (the medullary sheath) which in drying becomes black and pruinose; the outer layer or exorhiza splits at each end longitudinally for some length into four equal divisions, thus leaving both ends of the enclosed inner layer (the neorhiza) somewhat exposed. The embryo, as thus described, is shown in Plate VII. figs. 3, 4, 5, taken from Loureiro’s own specimen, now preserved in the British Museum, of his Eugenia acutangula (non Linn.) Its structure is evidently analogous to that of the embryo in Agasta, Butonica, Stravadium, and Careya, only more pronounced and better corresponding with Gaertner's plate 101 and with Dr. Thomson's instructive observations. То this genus I have referred the Careya macrostachya of Jack, and the Careya pendula of Griffith. DoxomMa, nob. Calycis adnati limbus breviter cupuliformis, margine in sepala divisus; sepala 4, squalia, rotundata, subcoriacea, ætivatione subimbricata, persistentia. Petala 4, oblonga, calyce longiora, crassiuscula, marginibus membranaceis retroflexis, unguibus tubo staminigero agglutinatis et cum illo caduca. Stamina numerosissima, multiseriata, imo in tubum subbrevem cylindricum monadelpha, seriebus interioribus вере brevioribus, setiformibus, et anantheris. Stylus illis equilongus, tenuiter subulatus, paullo curvatus. Stigma simplex. Discus plane annularis, margine exteriore tubum staminige- rum sustinens, intus in prominentiam erectam orelato expansus. Ovarium inferum, turbinatum, tetragonum, 4-loculare, vertice intra discum concavum ; ovula in quoque loculo 2, ab apice funiculis suspensa. Fructus cylindrice oblongus, subtetragonus, angulis costatis aut anguste alatis, unilocularis, monospermus ; pericarpium subcrassum, fibroso-coriaceum, calyce coronatum : Semen loeulum im- plens, funiculo suspensum ; testa tenuiter coriacea, pericarpio subadhzerens : embryo conformis, homo- neus, dure eburneus, mesopodus, ezorhiza а neorhiza soluta, ad utramque extremitatem 4-fissa et hiante, intus pulvere nigro tecta, neorhiza in germinatione,ut in Butonica, utrinque propullante. $ 1. Petioli elongati. 1. Doxomma PENDULUM, nob.: Careya pendula, Griffith, Notulæ, pars iv. p. 661, tab. 634 a. fig. 1 ad 10 select.: arbuscula, ramulis subtenuibus, levibus, rubi- dulis: foliis in apice ramulorum confertis, longe lanceolatis, utrinque valde acutis, marginibus cartilagineis subrevolutis obsoletissime serratis, chartaceis, supra pallide ігі ав, opacis, nervis tenuibus patentim divaricatis marginem versus nexis % 100 MR. J. MIERS ОМ THE ВАКВГУСТОМТАСЕЖ. prominulis valde reticulatis, subtus paullo pallidioribus, costa striata valde pro- minente, nervis stramineis prominentibus; petiolo tereti, imo crassiore, limbo 8plo breviore: racemo terminali, multifloro, nutante ; rachi valida, imo crassiore, angu- lato-suleata, colorata; floribus majuseulis, subapproximatis, subsessilibus vel bre- vissime pedicellatis, bracteolis minutis subulatis imo donatis; calyce cupulari, fere ad basin in lobos 4 subimbricatos rotundatos согіасеоѕ marginibus membranaceis ciliato-denticulatis diviso; petalis 4, expansis, crassiusculis, profunde rubris, elongato-ovatis, calycis 3pla longitudine, marginibus revolutis, unguibus ad tubum staminigerum affixis еб cum illo caducis; staminibus numerosissimis, albis, quam petala longioribus, pluriseriatis, interioribus sterilibus ; disco annulato, margine ex- teriore tubum staminiferum erectum fuleiente, intus in urceolum verticem ovarii con- cavum cingentem producto; ovario subcylindrico, tetragono, ovulis 3 in quoque loculo funiculis brevibus suspensis: fructu (vix maturo) cylindrice oblongo, apice subconstricto et calyce coronato, subtetragono, levi, rubidulo, pericarpio crasso, abortu l-loculari et monospermo; semine ovato, cum ovulis abortivis pendulo, raphe lineari cum funieulo crassiusculo continua, testa reticulata, imo chalaza sig- nato. Inregione Malayana et Borneo: v. s. in hb. Hook. Mergui к iffith), Mergui (Parish cum icone), Sarawak (Beccari 3255). А. low tree, growing in woods on hill-sides. А good floral idv is given both by Griffith and Parish. The branchlets are 2 in. thick; the leaves are 11-14 in. long, 21-3 in. broad, on petioles 11—13 in. long, with about 15 pairs of nerves, and others shorter and intermediate. Тһе raceme is 21 feet long, with a stout sulcated rachis 3-5 lines thick; flowers 1—5 in. apart; calycine lobes 1 in. long and broad; petals 14 in. long, 1 in. broad; stamens 2 in. long ; fruit 22 in. long, 11 lines broad ; seed 2 in. long, $ in. broad. Griffith, in p. 662, refers to a drawing of this in his tab. 634.4; but there we find à jumble of figures, without numbers, belonging to Careya and Cornus, 15 figures be- longing to the former, the remainder to the latter genus. Of the former I take ten to represent the seed of this species, the two top figures on the left hand showing them in their natural size as above described; another figure shows it magnified to double that size ; and seven others exhibit the same, with the embryo before and after it commences germination, showing at the plumular end a curving subulate shoot, furnished at its base with imbricating scales. The five larger figures before mentioned seem to belong to his Barringtonia conoidea, for which there was no room in his Plate 615. This intermixture of figures without explanation is attributable to the editor of Griffith's ‘Notule.’ A drawing of this species (with its analyses) is shown in Plate XV. fig. 9 et seq. 2. DoxoMMA CYLINDROSTACHYA, nob.: Barringtonia cylindrostachya, Griffith, Notule, part iv. p. 655: ramulis crassis, striatis, opacis : foliis subsparsis, longissime petio- latis, lanceolatis vel lanceolato-oblongis, apice breviter acuminatis, imo cuneatis, margine cartilagineo subrevoluto obsoletissime serratis, chartaceis, supra pallide viridibus, opacis, ad nervos suleatis, nervis plurimis patenti-divaricatis, intra mar- ginem nexis, transversim reticulatis, subtus flavidioribus, opacis, nervis valde prominentibus, costa prominente; petiolo tereti, striolato, pallide brunneo, imo MR. J. MIERS ON THE BARRINGTONIACEÆ. 101 valde incrassato, limbo 4plo breviore: racemo terminali, folio subzequilongo ; rachi erassa, cylindrica, densiflora, e floribus caducis creberrime cicatricata; floribus sessilibus ; calyce pocilliformi crassiusculo, extus brevissime puberulo, fere ad basin in lobos 4 rotundatos diviso; petalis 4, obovatis, crassiusculis, calyce triplo lon- gioribus ; staminibus pluriseriatis, imo in tubulum longiusculum monadelphis : disco annulato, margine externo tubum staminigerum fulciente, intus in prominentiam erectam crenulatam verticem concavum cingentem producto ; ovario turbinato, acute tetragono, angulis costatis, subeorrugulato, 4-loculari, ovulis in quoque loculo 4—5, axi affixis. In Birma: v. s. in hb. Hook. prope Avam (Griffith 2421), Malaya ( Maingay 163). This differs from the preceding species in its much longer petioles, and racemes with more crowded flowers: the axils of the leaves are 1—3 in. apart ; the leaves are 73—14 in. long, 23—81 in. broad, on a petiole 13—21 in. long, aid with about 30 pairs of nerves; the raceme is 1 foot long, or longer; the rachis 8 lines thick, dark lilae, closely beset with smallish cicatrices of the fallen flowers; the ovary is 4 lines long; the calyx 3 lines, the petals 9 lines long, 4 lines broad ; the stamens about 2 inches long ; the style 3 in. long, of a deep red colour: 4 оғ 5 ovules in each cell, upper ones ascending, the lower sub- pendulous. З. DOXOMMA COCHINCHINENSE, nob.: Еидетіа acutangula, Lour. (non Linn.) Coch. i. 375 (excl. synon.) : Stravadium cochinchinense, Bl. in V. Houtte, Flor. Serr. vii. p. 24: arbor, ramulis patentibus, fistulosis, brunneis, angulato-sulcatis, apice confertim foliiferis : foliis lanceolato-oblongis, apice in acumen angustum obtusulum constrictis, imo acute cuneatis, ad margines cartilagineos revolutos repandos obsolete serratis, subeoriaceis, supra pallidissime viridibus, nervis patentim divaricatis juxta mar- ginem nexis prominulis, valde reticulatis, subtus ochraceo-flavescentibus, valde opacis, nervis venisque prominentibus; petiolo longissimo, limbi fere longitudine, tereti, pallide opaco: racemo (sec. cl. Lour.) longissimo, pendulo, simplici ; sepalis 4, rotundatis; petalis 4, subparvis, subrotundis, concavis, conniventibus ; ovario infero, rotunde turbinato: fructu cylindrice oblongo, utrinque subaeuto, ealyce coronato, tetragono, angulis acutis; pericarpio corticoso, subrugoso, flavide fusco; semine unico, magno, oblongo (farinulento dum vivo 2), demum durissimo. Іп Cochinchina, in sylvis: v. pl. s. іп hb. Mus. Brit. loc. cit. роне, specim. typ.); v. fruct. i Mus. Brit. (Loureiro). From the typical specimen and fruit, this species may now be distinguished from the many heterogeneous plants erroneously referred to it. Although Loureiro recorded it as the Eugenia acutangula of Linnzus, he was aware of its nearer affinity to the Butonica rubra of Rumphius. Не described it as a large tree with spreading branches. Its leaves are 4-10 in. long, 2-23 in. broad, on petioles 3-7 in. long; they have about 14 pairs of patent fine nerves; the terminal raceme is often 2 feet long. The size of the flowers is not given; but the petals are stated to be small. The typical specimen of the fruit is 3 in, long, 11 in. broad, оп a short stout pedicel; it is narrowed at each extremity, 4 lines broad at the throat; where it is crowned by the calycine limb 8 lines long. Its Shape is somewhat fusiform and striolated, with 4 longitudinal wings, 11 line broad, and SECOND SERIES.— BOTANY, VOL. I. P 102 MR. J. MIERS ОМ THE BARRINGTONIACEÆ. wavy on the margin; the pericarp is fibrous, 1 line thick, and contains a single sus- pended всей, which fills the cell, and which has been fully described in the generic diagnosis (ante p. 99). А drawing of the icaf and of the carpical structure of this species is shown in Tab. XVI. 4. DoxoMMA ACUMINATUM, nob.: Stravadiwm acuminatum, Bl. in Van Houtte, Flor. Serr. vii. p. 24; Wall. Cat. 3636: Barringtonia (Careya) rosea, Wall. Cat. 3636: Barringtonia acuminata, Korth. in Kruidk. Ned. Ind. Archip. p. 206; Miq. Fl. Ned. Ind. i. 490: foliis lanceolatis vel oblongo-lanceolatis, apice in acumen breve attenuatis, imo sensim acutis, marginibus tenuiter cartilagineis, obsole- tissime serratis, tenuiter chartaceis, supra pallide viridibus, opacis, nervis tenuis- simis, divarieatis, reticulatis, subtus concoloribus aut paullo pallidioribus, nervis venisque prominulis, costa tenui, striolata; petiolo tenuissimo, semitereti, imo gihbose inerassato, limbo 6plo breviore: racemis spicatifloris; rachi cylindrica, valida, sub medium e floribus caducis crebre cicatricata ; floribus subparvis, subapproximatis, sessilibus; calycis limbo cupuliformi, coriaceo, ad basin in sepala 4 diviso, sepalis ferrugineo-pruinosis; petalis 4, roseis, calyce paullo longioribus ; staminibus nume- rosissimis, imo in tubum monadelphis; disco annulari, margine externo tubum sta- miniferum fulciente, intus in prominentiam fornicatam verticem concavum ambiente producto; ovario quadratim turbinato, ferrugineo-pruinoso, angulis cum lobis calycinis alternis, alatis, suberenatis, 4-loculari, ovulis 3-4 in quoque loculo suspensis : fructu elongato-ellipsoideo, utrinque obtuso, levi, subtetragono. In Malacca et Borneo: v. в. іп hb. Soc. Linn. Chappadong in ripis (Wall. Cat. 3636) ; fruct. non vidi. Wallich’s specimen from the province of Martaban agrees with Blume's diagnosis of а plant first mentioned, by name only, by Korthals, and to which he assigned the locality of Borneo. Wallich at first gave it the name of Barringtonia ( Careya) rosea, but afterwards changed it to Barringtonia acuminata, considering it to be the same as Korthals's, perhaps ; and if so, its range must be considerable. In Wallich's specimens the leaves are 51-8 in. long, 15-23 in. broad, on petioles 1-2 in. long; Blume gives the dimensions as 84-H in. long, 195-35 in. broad, on petioles 2-23 in. long, from a plant collected by Korthals in the woods of Pamattan, in Borneo. Тһе raceme in Wallich's specimen is above 9 in. long, 3 lines thick, straight; the sepals are 2 lines long, 3 lines broad ; the 4 angles of the ovary are winged and undulately crenated; the inner margin of the disk is acute and elevated, within whieh the vertex is deeply concave. 5. DoxomMa SARCOSTACHYS, nob.: Stravadium Sarcostach; уз, Blume, in Van Houtte, Fl. бетт. vii. p. 24: Barringtonia sarcostachys, Miq. in Fl. Ned. Ind. 90: foliis lanceolato- oblongis, apice obtusis vel acutis, imo cuneatis, integerrimis, nervosis ; petiolo tereti, т limbo 4plo breviore: racemo spicato, rachi cylindrica, crassa, carnosa, recta: fructu | -subtetragono.. In ins. Sumatra: non vidi. | | Blume? s short diagnosis is copied by Miquel, who adds, this species is very different from D. sumatrana. Both state that the leaves are on in. va 23-34 in. broad, on petioles хам от sometimes 31 in. кмк UNT T^ id MR. J. MIERS ON THE BARRINGTONIACEA. 105 6. DoxoMMA SUMATRANA, nob.: Barringtonia sumatrana, Miq. in Fl. Ned. Ind. i. Suppl. р. 315: foliis elliptico-oblongis, apice in acumen obtusulum vel acutum subito con- strictis, imo cuneatis, subcoriaceis, crenato-serrulatis, dentibus obtusis, заре mucro- niformibus, supra lucidis, costa nervisque subpatulis prominulis, subtus nervis parum prominulis; petiolo longiusculo, supra subplano, marginibus acutis: racemo cylindraceo, terminali, erecto : fructu oblongo, acute tetragono, apice sepalis 4 obtuse ovatis coronato, levi, pallido. In Sumatra: v. s. in hb. Hook. Sarawak (Beccari 851, 1554). | ; A species, according to Miquel, distinet from D. Sarcostachys, though closely allied to it; the leaves are said to be 3-83 in. long, 13-28 in. broad, on petioles 2-23 in. long: no other particulars are given. I have referred here Beccari's specimen 851, from Sarawak, which agrees well with Miquel’s character as to the leaves, but differs in its shorter petioles : here the leaves are elliptic-oblong, with a narrow acumen at its apex near $ іп. long; they are gradually subcuneate to the base, crenately serrulate on the МАБ, the teeth sometimes nearly obsolete and mucroniform, they аге from 22-10 in. long, 13-21 in. broad, on semiterete petioles 2-14 in. long, with slender patent arcuated nerves ; and the rachis is stout, more than 7 in. long (broken off at the summit) ; the sepals rounded, somewhat acute, smooth, fleshy, entire, 4 lines long and broad, united in a short сар at their base; the petals are obovate, rounded ; the ovary is turbinate, rugulose, and sessile. In Beccari's 1554 the leaves are more lanceolate and paler. 7. ЙохоммА NEO-CALEDONICUM, nob.: Barringtonia neocaledonica, Vieill. in Proc. Soc. Linn. Normand. viii. p. 10: arbor excelsa, ramis erectis; ramulis subtenuibus, арісе crebre foliiferis, inferne foliis delapsis cicatricatis: foliis lanceolato-oblongis, utrinque aeutis, ad marginem subrevolutum cartilagineum obsoletissime serratis, supra læte viridibus, nervis tenuibus adscendenti-divarieatis prominulis, valde reti- culatis, subtus flavide pallidioribus, glaucescenter opacis, nervis venisque prominenti- bus; petiolo subtenui, tereti, limbo 15-plo breviore: racemo terminali, pendulo, spicato, rachi cylindrica, erassiuseula, imo compressa, albide glauca, ereberrime multiflora ; floribus amplis, sessilibus, imo setaceo-bracteolatis ; calyce cupuliformi, fere ad basin in sepala 4 ovata obtusa glabra subimbricata diviso, sepalis coriaceis, margine late submembranaceis, ciliato-denticulatis, extus albide seabridulis ; petalis 4, oblongo- ovatis, erassis, albide roseis, sepalis 4plo longioribus; staminibus numerosissimis, pluriseriatis, imo in tubum brevem monadelphis, seriebus interioribus sæpe anan- theris; disco plano, margine exteriore tubum staminigerum fulciente, intus in urce- olum latum erectum producto, vertice hine cirea stylum concavo; ovario infero, ob- conico, ruguloso, alatim tetragono, alis rectis, angustis, 4-loculari, ovulis in quoque loeulo 4, suspensis: fructu oblongo, cylindrieo, calyce coronato. In Nova Cale- donia, in sylvis humilioribus: v. pl. s. in hb. Hook. Gatasse, Wagap ard 2630) ; fructum non vidi. _ tall tree, 65-80 feet high, with leaves about 1 in. apart, 8-83 in. long, 28—28 in. broad, 9n petioles 6-8 lines long: the raceme is more tkan 5 in. long, with a cylindrical straight rachis 3 lines thick, studded in many places with close cicatrices of the fallen flowers : : P2 104 МВ. J. MIERS ОМ THE BARRINGTONIACE. the handsome flowers are rather crowded and sessile; the calycine sepals аге 3 lines long; the petals 12 lines long, 9 lines broad. The fruit is said to be 35 in. long. 8. DoxoMMA RIGIDUM, nob.: ramulis crassis, sulcato-angulatis, dealbatis: foliis majus- culis, lanceolato-oblongis, apice sensim acuminatis, imo gradatim acutis, marginibus subrevolutis fere integris et punctis nigris subremotis signatis, rigide coriaceis, supra intense viridibus, opacis, obsolete corrugulatis, ad costam suleatis, nervis divaricatis semiimmersis, subtus pallidioribus, costa valida, imo sensim incrassata, striolata, fusca, nervisque fuscis prominentibus, marginem versus nexis, venis transversim reticulatis; petiolo valido, imo crassiore, semitereti, fusco, supra sulcato, subtus striolato, limbo 20plo breviore: racemis axillaribus, fructiferis, superioribus folio subzquilongis, inferioribus multo brevioribus, rachi rigida, subflexuosa, alternatim prominenter nodosa : drupis in nodis sessilibus, quadrato-cylindricis, angulis rotunda- tis, utrinque subtruncatis, lobis calycinis 4, parvis, erectis, rotundatis, concavis, eoro- natis; pericarpio in fruetu immaturo crasse coriaceo, pallido, extus nitido, intus crasse coriaceo, 4-loculari, loculis 3 subabortivis, uno majore, ovulis plerumque mar- cescentibus. In Malaya: v. s. in hb. Hook. (Maingay 767, 2496). A peculiar species, with stout branchlets 5 lines thick, and axils 4-2 in. apart; leaves rigid, 94-114} in. long, 35—83 in. broad, on stout rigid curving petioles 14-14 in. long; raceme in the superior axil above 12 in. long, in the lower axils 2-3 in. long, all fructi- ferous, the flowers having fallen off: the rachis is stout, suberect, angulated, with alter- nate nodose axils 2—4 lines apart, at first little prominent, but in fruit the nodes are thick and prominent ; the fruit, sessile upon them, is, in the immature state, 1 in. long, 4 lines broad. 9. DoxoMMA MACROSTACHYUM, nob.: Careya macrostachya, Jack in Cale. Journ. iv. 335; in Hook. Bot. Misc. ii. 88: arbor ramulis lzvibus, cinereis: foliis sparsis, obovato- vel oblongo-ovatis, apice obtuso in acumen breve constrictis, imo sensim cuneato-angustatis, tenuiter chartaceis, margine tenui cartilagineo subrevoluto obsolete serrulatis, supra levibus, læte viridibus, nervis rubellis tenuibus semi- immersis, subtus pallide flaviusculis, costa nervis venisque valde reticulatis promi- nulis; petiolo subtenui, imo incrassato, supra canaliculato, limbo 5-Тріо breviore : racemis dependentibus, rachi crassissima, basi erecta, ereberrime multiflora ; floribus speciosis, sessilibus, spiraliter enatis, ebracteatis; sepalis 4, rotundatis, glabris, denti- culatis, purpurascentibus; petalis 4, triplo longioribus, obtuse ovatis, unguieulatis, rubris; staminibus albis, nuimerosissimis, imo monadelphis, petalis paullo longi- oribus; disco plano annulato, margine externo tubum staminigerum fulciente, interno in urceolum elevatum producto, urceolo intus rubro-striato, ore lato integro et flavido, profunde circa stylum ambiente; ovario turbinato, infero, 4-loculari, ovulis 4 in quoque loculo suspensis; fructu (вес. cl. Jack) baccato vel pomiformi. In Penang et Borneo: v. pl. s. in herb. Hook. Sarawak (Веесат 1535) ; fruct. non vidi. | A tree, with grey bark and smooth branehes, found by Jack at Pulo Penang, who MR. J. MIERS ON THE BARRINGTONIACE Ж. 105 remarks that in its expansive inflorescence, which is very remarkable, it is unlike any known species of Careya. Т have nowhere met with any plant which agrees with Dr. Jack's description, except that above quoted, found in Borneo by Beccari, the distances between the two places being only 600 miles of intervening sea. Jack's account differs in no respect from Beccari’s plant except in the character of his nectarium, which is said to be hypocrateriform ; but this cannot be; it is certainly a mistake, caused by the editor of his notes, in substituting that word for crateriform. With this correction, Jack’s account admirably conforms with the structure seen in Beccari's plant; we ean therefore have little doubt of the specific identity of the two plants. Jack does not give the size of the leaves, though in form both are alike: in Beccari’s specimen they are 21-53 in. long, 13—31 in. broad, on slender petioles, thickened at their base, 7-9 lines long. The rachis of the inflorescence, according to Jack, is thick, massive, cylin- drieal, 8-10 in. long, bearing crowded sessile rather large handsome flowers. In Beccari’s specimen the turbinated ovary, 3 lines long, is 12 line thick at base, 4 lines above; the sepals are 3 lines long and broad ; the petals, thick and fleshy, are 10 lines long, 6 lines broad; the stamens, 11 lines long, are seated on a monadelphous tube 1 line long; the elevated expansion of the inner margin of the disk (nectarium, Jack) is two lines higher than the disk, but much deeper within and around the style, its entire margin being 21 lines in diameter: the structure of the ovary is as I have described it. Wallich wrongly referred to this species his S/ravadiwm angustatum ; Roxburgh was alike in error with respect to his Barringtonia racemosa, and Griffith also to his Careya pendula, as well as Miquel in regard to his S/ravadiwm costatum. АП these surmises are without any evidence to support them. ' $ 2. Реной subbreves. 10. Похомма ANGUSTATUM, nob.: Stravadium angustatum, Wall. т Cat. 3637: ramulis erassiusculis, angulato-suleatis, pallide brunneis, fistulosis: foliis approximatis, lan- ceolato-oblongis, apice in acumen acutum attenuatis, a medio gradatim cuneatis ad imum angustum subito rotundatis, margine cartilagineo subrevoluto crenatis, et obsoletissime serrulatis, flaccide coriaceis, supra pallidissime glaucis, opacis, nervis tenuibus divaricatis prominulis, reticulatis, subtus fere concoloribus, nervis venisque reticulatis prominentibus, costa striolata paullo prominente ; petiolo compresso, lati- usculo, supra plano, subtus eonvexo et striolato, pruinoso, limbo 80-90plo breviore : racemo terminali, stricto, folio multo breviore; rachi cylindrica, undique ereberrime florifera; floribus pro genere parvis, sessilibus, bracteolatis ; calyce cupuliformi, fere ad basin in sepala rotundata coriacea margine ciliata diviso; petalis 4, sepalis 2plo longioribus, membranaceis ; staminibus numerosissimis, imo in tubulum erec- tum monadelphis; disco annulari, extus tubum staminiferum fulciente, теріне interno in prominentiam brevem vertice concavo ambiente producto ; ovario infero, quadrato, angulis alis rigide coriaceis crenato-undulatis munito, albide pulverulento, 4-loculari, ovulis 2-4 in quoque loculo suspensis. In Tenasserim: г. s. in hb. Soc. Linn. et Hook. Amherst (Wall. Cat. 3637 cum not; an Careya macrostachya, Jack ?) ; Tenasserim (Wall. Cat. 3631, Helfors). 106 MR. J. MIERS ON THE BARRINGTONIACE./E. This plant agrees in its general characters with the preceding species, and differs from Jack's plant in its more lanceolate and very pallid leaves, on almost obsolete flat petioles, and its spicated raceme, which is not massive, and its ovary with crenulately undulated wings. Тһе leaves are 9-17 in. long, 8-42 in. broad, on petioles 2 lines long, 2 lines broad; the portion of the raceme in the specimen is 6 in. long, 2% lines thick, closely beset with the rather small cicatrices of the fallen flowers: the calyx is 3 lines long; the ovary of the same length, with broad crenulated wiuged angles; the petals are 6 lines long, 3 lines broad. 11. Doxomma Увтьзи, пор. : Barringtonia Vriesii, Те. & Benn. Nat. Tijd. Ned. Ind. 11. 308; Kruidk. Arch. iii. 411; Miq. Fl. Ned. Ind. 1. 491 e£ 1087; Walp. Ann. iv. 852: arbor mediocris, ramulis roridis, hepatice viridibus: foliis oblongo-lanceolatis, apice anguste acuminatis, imo euneatim acutis, marginibus obsolete serrulatis, utrinque glabris, supra lucide viridibus, costa plana; petiolo hepatico, supra plano: racemo terminali, rachi valida, erecta, folio longiore, spicatim densiflora, pedicellis roridis, imo bibraeteolatis; petalis speciosis, albido viridulis: fructu majusculo, oblongo, utrinque sensim attenuato, tetragono, hepatice viridi. Java, in prov. Bantam (won vidi). : This is said to be a handsome tree, not very high, with wide-spreading branches, grow- ing in moist places: its leaves are 4—55 in. long, 14-2 in. broad, with liver-coloured petioles : the raceme is erect, 4—7 in. long, with numerous very handsome approximated flowers, having whitish-green petals, which, with the stamens, fall off in great numbers, and appear like a party-coloured carpet beneath the tree. Miquel, in his diagnosis, which is an exact copy of that of the Dutch authors, by mistake, says “ bace:e verru- созге,” not observing the typical error in the original, * drupis verrucibus " for ** drupis viridibus," the latter being so described in their annotation. Miquel admits that he never saw the plant, not even a dried specimen of it, nor its fruit. He adds that the genus to which it belongs is doubtful, because of its bibracteolated flowers, and suggests, without any reason, that it may һе the Planchonia Timoriensis, Bl. But subsequently he again falls into error, stating positively that it is identical with the Barringtonia excelsa, Bl.; but this cannot be admitted, because the latter is a very lofty tree, with а remarkable and different inflorescence; and, notwithstanding this, he again suggests that it may be the type of a new genus. І have no doubt that it belongs to Doromma, of which it presents all the characters, especially in its stout erect rachis densely covered with numerous handsome bracteolated flowers. Teijmann and Bennindijn describe the fruit as a little longer than a goose’s egg, pointed at both extremities, and liver-coloured. 12. Йохомма MAGNIFICUM, nob.: ramulis validissimis, summo breviter constrictis et setaceo-ramentosis, dein crassioribus et cicatricibus petiolorum magnis creberrime 2% signatis : foliis majusculis, elongato- et late oblongis, apice in acumen brevissimum obtusum constrictis, sub medium euneato-angustatis et circa petiolum subito rotundi- usculis, in marginibus incrassatis, obsolete serrulatis, subcoriaceis, supra pallidissime viridibus, opacis, nervis remotis areuatim adscendentibus tenuibus vix prominulis | MR. J. MIERS ON THE BARRINGTONIACE.JE. 107 reticulatis, costa valida, imo sensim crassiore, subtus palidissime lutescentibus, nervis costatis, venis prominulis, costa convexa, valde prominente; petiolo latissimo semitereti, limbo 30plo breviore: racemo terminali, longissimo, mox pendulo; rachi cylindrica, crassissima, profunde striata, suberebriflora; floribus in paribus super- positis et sessilibus, imo bractea majuscula lanceolata decidua munitis ; calyce cupu- liformi, fere ad basin in lobos 4 imbricato-oblongos concavos carnosulos diviso, extus brevissime puberulo; petalis 4, oblongis, marginibus late revolutis, calyce 3plo lon- gioribus, viridi-albidis, expansis ; staminibus numerosissimis, imo in tubulum erectum monadelphis et cum petalis imo agglutinatis simul caducis ; disco annulato, margine externo tubum staminigerum fulciente, intus in prominentiam elevatam styli basin cingentem producto; ovario subcylindrico, tetragono, obsolete puberulo, 4-loculari : fructu immaturo subparvo oblongo-ovato, utrinque obtuso, calyce coronato, quadrato, angulis subalatis, abortu l-loculari et monospermo. Іп Tenasserim: v. s. in hb. Hook. Тауоу (Parish). А tree 30—40 feet high, with a straight trunk and a fine head of spreading branches, called by the natives kyai-gyce; it is also found at Moulmcin. It forms a species quite distinet from most of the preceding, differing in its much broader leaves, on short petioles, and in its flowers arranged in pairs, and its fruit with winged margins. Its leaves, very crowded at the extremity of the branches, are 16 in. long, 6 in. broad, on a petiole 6 lines long, 4$ lines broad, with about 13 pairs of very spreading nerves; raceme upwards of 3 feet long, rachis 6 lines thick below; pedicels 1 line long; calyx 4 lines long; petals 10 lines long, 6 lines broad; stamens above an inch long; fruit (perhaps immature) 11 lines long, 8 lines broad, quadrate, with winged margins, its single seed is suspended, 5 lines long. 8. PETERSIA. This is a singular genus, proposed by Dr. Welwitsch for a plant discovered by him in Angola, a tree of large size, with an expanded сотозе head: it has terminal pendent racemes, оп a rather thin rachis, bearing many small flowers, each upon a long slender pedicel, which is bibracteolated in the middle; the calycine limb consists of 4 small rounded erect sepals, distinct in the bud; 4 obtuse petals, fixed by their claws to the staminiferous tube; an inferior turbinate 4-winged ovary, which is 4-locular, with several ovules suspended in the summit of each cell; the fruit is of a narrow linear cylindrical shape, coarsely hispid, with 4 very large extremely broad membranaceous wings. By abortion it is unilocular, with few linear compressed suspended seeds. РЕТЕЕ8ТА, Welwitsch. Calycis adnati limbus in sepala divisus. Sepala 4, rotundata, erecta, ciliata, in :stivatione subimbricata, persistentia. Petala 4, cuneato-oblonga, membranacea, calyce longiora, unguibus ad tubum stami- nigerum affixa, et cum illo caduca. Stamina numerosissima, multiseriata, imo in вара brevem eylindricum monadelpha : filamenta tenuia erispatim flexuosa ; discus anguste snpularis, externe tubum staminigerum fulciens, interne in urceolum brevem productus ; stylus tenuis, longitudine sta- _ minum; stigma parvum, clavatum. Ovarium inferum lineari-turbinatum, 4-alatum, alis imo pue . lete 4-lobatis in pedicellum deliquescentibus, apice rotundatis, sepalis alternis, 4-loeulare: ovu/a in 108 MR. 7. MIERS ON THE BARRINGTONIACE Ж. quoque loculo plurima, axi summo suspensa. Fructus lineari-cylindricus, latissime 4-alatus, айз magnis, latissimis, apice profunde semicordatis, membranaceis, crebre parallele et horizontaliter nervosis; pericarpium lineare, tenuiter fibroso-coriaceum, abortu 1-1осшаге ; semina pauca, lineari- oblonga, сошргезза, subimbricatim ab apice suspensa. Arbor Angolensis, grandis, frondosa: folia alterna, elliptico-oblonga, utrinque acuta, breviter petiolata ; racemus terminalis, brevis, pauciflorus ; flores alternatim pedicellati, inconspicui ; fructus singulariter late 4-alatus, majusculus. PETERSIA AFRICANA, Welwitsch, Hook. & Benth. Gen. Pl. i. 721; Oliv. Fl. Afr. Trop. ii. p. 439: procera, coma magna dilatata, apice depressa, ramis dependentibus: foliis epunctatis, oblongo-ellipticis, apice in acumen acutum subito constrictis, imo sensim cuneatis, marginibus obsoletissime serratis, subcoriaceis, supra sublucidis, intense viridibus, nervis tenuibus divaricatis marginem versus nexis vix prominulis, subtus pallidioribus, opacis, nervis. prominulis in axillis barbatis; petiolo tereti, margi- nato, limbo 12plo breviore: racemis in ramulis novellis terminalibus aut sub- axilaribus, folio brevioribus, paucifloris; floribus subparvis, longe pedicellatis ; pedicellis tenuibus, elongatis, medio bibracteolatis, bracteolis parvis, lanceolatis, oppositis, puberulis; calycis limbo in sepala 4 semiovata erecta puberula ciliata diviso; petalis 4, rotundo-oblongis, glabris, unguibus tubo staminigero agglutinatis ; staminibus numerosissimis, petalis longioribus, imo in tubum brevem erectum monadelphis; ovario infero, turbinato, 4-alato, alis latiusculis, ciliatis, 4-loeulari, ovulis in quoque loculo plurimis, pluriseriatis, axi centrali affixis : fructu lineari, tetra- gono, latissime 4-alato, alis membranaceis, glabris, parallele nervosis, superne pro- funde cordatis; pericarpio extus setis rigidis hispido, textura fibroso; endocarpio tenuiter coriaceo, aborto 1-loculari, axi e loculis abortivis hinc parietali; seminibus 4-5, linearibus, compressis, imbricatis, funiculis ab apice suspensis. In Africa occi- dentali, in сопуа из humidioribus Monte de Quetecati, distr. Cal. alto, regno Ango- lensi: v. s.in hb. Hook. (W'elwitsch). An immense tree, 50 feet high, its trunk 4 feet in diameter to a height of 20 feet, when it throws out many spreading and somewhat pendent branches, forming a head rather depressed at its summit. The leaves are 31—61 in. long, 12-21 in. broad, on petioles 3—6 lines long, having 12-14 pairs of nerves, with a tuft of hairs at their origin: it is stated (Hook. & Benth. Gen. Plant. i. p. 721) that its leaves are pellucidly punctate. I contested this point with the late Dr. Welwitsch, showing him that in all the younger leaves, and in nine cases out of ten, no pellucid dots are visible, and that where they were present, this was due to round globules of transparent matter deposited by insects in the parenchyma, sometimes in clefts made by them, sometimes appearing like raised glands. The terminal raceme is 3 in. long, on a rather slender rachis; the pedicels are 6-9 lines long, bearing in their middle (where they are articulated) 2 opposite lanceolate bracts 13 line long; the ovary is 2 lines long, is broadly 4-winged; the sepals 1 line long; the petals 4 lines long, 2 lines broad ; the stamens 5 lines long ; the style 6 lines long. The fruit is altogether 22 in. long, 21 in. broad, the wings deeply cordate at the summit ; the central pericarpial portion begins at the deep sinus between the wings, extending down- MR. J. MIERS ON THE ВАККІМСТОМІАСЕ Д. 109 wards to the base; this portion is almost spinosely hirsute exteriorly: its texture con- sists of loose rigid longitudinal fibres without any agglutination; and, by abortion, it is only 1-celled, lined with a thinnish coriaceous еп4осагр, is 12 lines long and 2 lines in diameter; by the abortion of 3 of the cells, the central placental axis becomes parietal ; and upon it I observed four or five imbricated seeds, linear and much compressed, 3} lines long, $ line broad, suspended from the summit by a broad funicle 2 lines long; the embryo I could not make out. Dr. Welwitsch, in his note, quoted in the * Genera Plan- tarum,’ states he did not find perfect seeds; but the above details were obtained from a fruit, flowers, and plant kindly given to me by the late Dr. Welwitsch. А. drawing of this species (with its floral and carpological analyses) is shown in Tab. XVIII. figs. 16 to 26, 9. MEGADENDRON *. This genus is proposed for two handsome Java plants, very lofty trees, one of which is the Barringtonia macrocarpa, so well described by Hasskahl in 1844: it has large long leaves, and bears much resemblance in its handsome flowers, upon a long terminal raceme, to those of Doxomma; but it differs from that genus in the constantly variable number of its sepals and petals, in the inflected margins of its erect sepals (free even in :estivation), in the position of its ovules, and in its much more elongated fruit. MEGADENDRON, nob. Barringtonia, Hassk. et Bl. (non Forst.). Sepala 2, 3, 4, libera, acute oblonga, chartacea, parallele nervosa, erecta, jam inde in zstivatione margi- nibus involutis et ad se applicitis, persistentia. Petala 4, 5, 6, duplo majora, oblonga, sestivatione imbricata, dein expansa, lateribus retroflexis, unguibus tubo staminigero agglutinata, et cum illo caduca. Stamina numerosissima, in tubum brevem imo monadelpha ; filamenta filiformia, pluriseriata, petalis longiora, seriebus interioribus anantheris et capillaribus ; anther@ рагуге, 2-loculares. Discus epigynus, plane annularis, margine exteriore tubum monadelphum fulciens, interiore in urceolum brevem latum expansus. Stylus filiformis, longitudine staminum ; stigma parvum. Ovarium inferum, turbinatum, vertice intra urceolum et stylum concavo, 4-loculare, ovulis in quoque loculo 5-6 (raro 2-4) axi radiatim affixis. Fructus majusculus, elongato-cylindricus, utrinque obtusatus, calyce per- sistente styloque coronatus; pericarpium crassum, fibroso-coriaceum, abortu l-loculare et mono- spermum. Semen loculum implens, cylindricum, utrinque subattenuatum, obsolete tetragonum : testa fusca, submembranacea ; embryo exalbuminosus, amygdalinus, mesopodus, siccus exorhiza ad utramque extremitatem in fissuras 4 hiante. Arbores Timorenses et Javenses, excelse, ramose ; folia majuscula in apicibus ramorum congesta, ob- longa vel lanceolata, fere integra, petiolata ; flores speciosi, in racemo terminali sparse spicati. 1. MEGADENDRON MACROCARPUM, nob.: Barringtonia macrocarpa, Hassk. Diagn. Nov. p. 504; in Cat. Hort. Bogor. p. 263; in Bot. Zeit. xxvii. 598 (1844); Miq. Fl. Ned. Ind. i. 485 (1855): Barringtonia racemosa, var. elongata, Bl. in Van Houtte, Fl. Serr. vii. 24: arbor alta, ramis crassis, cinerascentibus : foliis in summo ramulorum crebre congestis, late oblongis, apice acuminatis, infra medium subito cuneatis, ad marginem subrevolutum crenato-serratis, coriaceis, glaberrimis, supra profunde viri- am * А péyas (magnus), dévepov (arbor). SECOND SERIES.— BOTANY, VOL. I. 9 110 MR. J. MIERS ON THE BARRINGTONIACEA. dibus, nervis paullo prominulis, costa plana, imo incrassata, subtus pallidioribus, nervis validis, albescentibus; petiolo brevi, crasso, semitereti, late marginato, atro- viridi, limbo 70plo breviore: racemo terminali, pendulo, folio :equilongo; floribus speciosis, sparse spicatis, pedicellatis ; sepalis 2, 3, 4, acutis, erectis, marginibus an- guste inflexis, chartaceis, parallele nervosis; petalis 4, 5, 6, obovatis, triplo longi- oribus, lateribus undulato-reflexis, albis ; staminibus disco styloque ut in char. gen. ; ovario infero, 4-loculari: fructu elongato-cylindrico, structura in char. gen. descripta. In Java et in insulis Sondaicis: v. s. in hb. Mus. Brit. Patjetan (Horsfield) ; in hb. Hook. Straits Sunda (Staunton); fruct. non vidi. А lofty tree, with spreading branches; its leaves are 10—18 in. long, 4—6 in. broad, on stout petioles 2 lines long and broad; raceme more than a foot long, with flowers 2-4 in. apart, 21 in. in diam. ; porous stoutish, nearly i in. long; ovary the length of the sepals, which are 4 ‚ long; petals 9-12 lines long; staminiferous tube y5 in. long; exterior filaments 11 in. TOM моё ones short, capillary without anthers; fruit 5 in. long, 14 in. broad at the angles, 1 in. broad at the sides; cell of pericarp 23 in. long; the seed fills the cell. Blume regarded this plant as a variety of Виютса racemosa, naming it elongata, оп account of its very long cylindrical fruit. Hasskahl gave copious and exact characters of this plant, which is almost the only species in the family well described; for Blume and Miquel have given only laconic characters, often contradic- tory and incomplete. Най they followed the example of Hasskahl, we should have obtained a much better knowledge of the family. А drawing of this species (with its floral and carpological analyses) is shown in Tab. XV. figs. 1 to 8. 2. MEGADENDRON PALLIDUM, nob.: foliis longe lanceolatis, apice sensim acuminatis, imo gradatim longe cuneatis, subsessilibus, sinuato-serratis vel marginibus sub- revolutis, hine punctis nigris remotis signatis, subcoriaceis, supra pallidissimis, nervis remotis subrecto-divergentibus marginem versus nexis, costa plana, immersa, imo gradatim valde incrassata, subtus fere concoloribus, costa nervisque prominentibus; petiolo vix ullo: racemo terminali, pendulo, rachi subtenui, compresse striata; floribus subsparsis, brevissime pedicellatis; sepalis 2—3, acutis, rectis, concavis, mar- ginibus anguste, involutis (petalis et staminibus lapsis); ovario infero, oblongo, ob- solete tetragono, superne disci urceolo stylum ambiente notato, 4-loculari, loculis 3 cum ovulis abortivis, in altero ovulo magno pendulo. In Java: v. s. i». hb. Mus. Brit. loc. cit. ( Horsfield). A species evidently belonging to this genus : its leaves are 191 in. long, 42 in. broad, quite sessile; the rachis of the raceme is not very stout, 15 in. long; pedicels 3—6 lines apart, less than 1 line long ; ovary grown after the fall of the petals is 2 in. long, erowned by 2, sometimes 3 sepals 5 lines long, 24-3 lines broad at the base. 3. MEGADENDRON ? AMBIGUUM, nob. : Barringtonia timorensis, Blume in Van Houtte, _ Flor. Serr. vii. p. 25; Мід. in Fl. Ind. Ned. i. p. 485: foliis oblongis vel elliptico- oblongis, apice acutis aut acuminatis, infra medium subcuneatis, basi subacutis MR. 7. MIERS ON THE BARRINGTONIACE Ж. 111 vel obtusis, obsolete serrulatis, nervis tenuibus, remotiusculis; petiolo limbo 30- 40plo breviore: racemo terminali, spicatifloro, longissimo, rachi crassiuscula, erecta, vel subnutante, glabra; floribus ignotis: fructu ovoideo, acute tetragono. In ins. Timor: non vidi. A species ambiguous in its relations. Blume places it next to his Barringlonia racemosa, which is numerous in synonyms, the chief of which is his var. elongata, the type of this genus. Miquel’s description of this species is а mere repetition of Blume's unsatisfactory diagnosis. It appears to me to belong to Megadendron, not only on account of the country of its origin, but its general habit, its long pointed cuneated leaves, suddenly rounded at base upon an extremely short petiole, a terminal long raceme, with an erect thickened rachis—characters peculiar to the genus to which I have assigned it. Blume and Miquel refer it to the Barringtonia acutangula, Spanaghoe, a species known by name only, without any given character, and evidently unknown to them. Blume states that the leaves in this species are 4-10 in. long, 11—41 in. broad, on petioles only 2-3 lines long. Хо other dimensions are given. 10. CHYDENANTHUS. This new genus! is founded upon a plant which differs from all others of the family in having a branching panieulated inflorescence, and which was originally described by Blume in 1826 under the name of Barringtonia or Stravadium excelsum. It is a native of Java, a lofty tree, with broadly spreading branches and long pointed leaves; the terminating panicle, longer than its large leaves, has alternate branches, presenting at rather short intervals alternate very prominent nodes, each having articulated upon it a sessile ebracteated flower of moderate size: the calyx is peculiar, of a cup-shaped form, with an almost entire margin, broadly open in the bud; it has 4 oblong petals, attached by their claws to the staminiferous tube as in the other genera; there is a peeuliarity in the staniens in having most of the filaments antheriferous, while the 2 or З inner rows are reduced to the form of slender barren pointed hairs: the inferior ovary is cylindrical, 8-grooved, 2-locular, with a single ovule erect in each cell : the fruit, according to Blume, is like that of Stravadium in form. CHYDENANTHUS, nob. Stravadium, Bl. ; Barringtonia, Miq. Calycis adnati limbus hemisphzrice cupuliformis, margine truncatus, vix lobatus, ciliolatus, subcoriaceus, persistens. Petala 4, obovata, quorum 2 paullo majora, calyce 4plo longiora, membranacea, вии ad tubum staminigerum а ха, et cum Шо caduca. Stamina numerosissima, multiseriata, жю 48 tubum cylindricum monadelpha, seriebus exterioribus fertilibus, interioribus multo brevioribus, tenuioribus, ad basin fere liberis, anantheris. Stylus tenuis, equilongus, Discus plane annularis, margine externo tubum staminigerum fulciens, intus in prominentiam оин expansus. Ova- rium inferum, semiglobosum, 2-loculare : ovula in quoque loculo solitaria, erecta. Fructus elon- gato-ellipsoideus, utrinque obtusus, tetragonus: cetera ignota. ; Arbor мүт wolis ; folia ИИ utrinque acuta, subobsolete serrata, breviter 5 From x2» (diffuse), áv8os (flos). zà 112 MR. J. MIERS ON THE BARRINGTONIACEÆ. petiolata ; inflorescentia terminalis vel subazillaris, singulariter. paniculata, rachi pubescente, ramis pedunculatis, alternatim longe nodosis ; flores sessiles, cum nodis articulati et plerumque tandem caduci, speciosi; fructus cylindrice oblongus, viz notus. 1. CHYDENANTHUS EXCELSUS, nob.: Barringtonia excelsa, Blume (non Benth.), Bijdr. 1097; Miq. Fl. Ned. Ind. i. p. 491: Stravadium excelsum, Bl. in DC. Prodr. iii. 289; Bl. in Van Houtte, Fl. Serr. vii. p. 24: arbor alta, ramulis pallide brunneis, opacis : foliis oblongo- vel lanceolato-ellipticis, apice subito et acute attenuatis, imo obtusis vel acutioribus, marginibus cartilagineis subrevolutis vix repando-serratis, chartaceis, supra pallide viridibus, valde opacis, sub lente granulosis, nervis tenuissimis sub- adscendentibus arcuatim nexis, semiimmersis, costa tenui, subtus pallide fulvescenti- bus, opacis, minute granulosis, costa nervis stramineis venisque transversim reticulatis subprominentibus; petiolo semitereti, corrugulato, limbo 20plo breviore: racemis axillaribus et simplicibus, folio dimidio brevioribus, vel terminalibus, longioribus et ramoso-paniculatis, fulvide tomentellis, ramis remotis aut verticellatim congestis, imo (ut in axillaribus) nudis, supra medium spicati-floris, nodis crassis projectis, alternis, erecto-divaricatis, singulis l-floris; floribus hine sessilibus et articulatis, plerumque caducis; calycis limbo cupuliformi, margine membranaceo, undulato, obsolete 4-lobato, griseo, puberulo; petalis 4, obovatis, quorum 2 paullo majoribus, membranaceis, minute puberulis, unguibus tubo staminigero affixis; staminibus numerosissimis, petalis longioribus, imo in tubum brevem monadelphis, seriebus interioribus eximie brevioribus, tenuioribus, anantheris, ad basin fere liberis; disco annulari, extus tubum staminigerum fulciente, margine interno іп tubulum crassum sulcatum stylum ambiente expanso; ovario infero, cylindrico, 8-sulcato, distincte 2-loculari, ovulo unico in quoque loculo e basi erecto: drupa elongato-ellipsoidea, utrinque obtusa, tetragona. Java, in maritimis: v. s. in hb. Mus. Brit. Prowats (Horsfield); in hort. Bogor cult. (Anderson); іп hb. Hook. Java (а cl. Miquel missa) : fructus non vidi. Miquel's specimen certainly belongs here, and, though named by him Barringtonia Vriesii, does not correspond with the species published under that name by Teysmann and Bennenden. Тһе details given by Blume are extremely short. Не describes it asa tree 80 feet high; its leaves are 5-72 in. long, 2-23 in. broad, on petioles 8-4 lines long. This is one of the rare instances hitherto known in the family where a paniculated inflorescence occurs; this is accompanied by other characters, showing it to form a distinct genus: the rhachis is 4-10 inches long, with 4 alternate branches somewhat distant, 2-3 in. long in one specimen, while in another these branches аге fasciculated or alternate upon а common peduncle 2 in. long; each branch has many projecting alternate thick nodules 1-2 lines long and 1-2 lines apart, each supporting a sessile flower artieulated upon it: the ovary is like a pedicel, 1 line long, } line thick; the cupular calyx is 2 lines long, 3 lines broad; the petals are И horizontally ex- tended, two of iini! 9 lines, the others 8 lines long, and 41 lines broad ; the outer stamens, 1 inch long, are united at their base into a tube 24 lines long, the inner ones are almost - free to the base, finely capillary, 3 lines long, and without anthers. MR. J. MIERS ON THE BARRINGTONIACES. 113 Blume makes this species a section (Anisostemon) of his Stravadium, distinguished by its peculiar inflorescence and its remarkably short interior stamens. Blume, in his Bijdragen, p. 1087, says that this plant is the same as the Barringtonia _ Friesii of Teysmann and Bennenden, which he suggests should be expunged. But Miquel does not accord in this view, and he quotes afresh the details given by the latter authors, that the inflorescence is an erect terminal raceme with dense flowers, seated upon sepa- rate pedicels bibracteolated at their base. He thinks it should be placed in a separate genus; but he confesses that he had not seen any specimen of it. I have referred it to the genus Doxomma. А drawing of this species (with its floral analysis) is shown in Tab. XVII. figs. 15 to 20. Fig. Ы 2. 3. 4. 5. 6. DESCRIPTION OF THE PLATES. PLATE X. À leaf of Barringtonia speciosa, from Forster's own specimen. . A flower of the same in bud, showing its entirely closed calyx. ‚ The flower expanded, copied from Forster's drawing. . The stamens of the same, combined in a monadelphous tube, which is cut open, and to which one of the petals is affixed by its claw: all nat. size. ‚ One of the anthers: much magnified. . The ovary, with the persistent calyx, the disk, with its inner nectarial expansion, and style : nat. size. . À longitudinal section of the same, showing two of its four cells, with its suspended ovules, the epigynous disk, which supports the staminal tube on its outer margin, is crenately lobed in the middle, and with its outer margin expanded into a long tube, dentated at its apex, and surrounding the lower part of style. ‚ The fruit crowned by the persistent calyx. : . The same, with half of the sarcocarp cut away transversely, to show the fibres surrounding the endocarp. . The endocarp, with its surrounding fibres. ‚ А transverse section of the 4-celled endocarp, with one of the enclosed seeds shown. . The embryo of the seed: all nat. size. Prats ХІ. The flower, in bud, of Ayasta splendida, showing its entirely closed calyx. The same, with the calyx splitting into 2 lobes by the expansion of the petals. A flower expanded upon its pedicel, which has a foliaceous bract in its middle, copied from Parkinson’s drawing. E A longitudinal section of the ovary, calyx, staminal tube, seated on the outer margin of the epigynous crenated disk, and style; two of its four cells are seen with the suspended ovules. À transverse section of its 4-celled ovary. À longitudmal section of the drupe, crowned by the persistent calyx and style, and its single large seed enclosed in its rather thick pericarp: all nat. size. оз It 10. 12. 12. 18. 14. MR. J. MIERS ON THE BARRINGTONIACE Ж. Ртлте XII. . Flower in bud of Agasta indica. . Calyx, opened, and ovary, from which the petals and stamens have fallen away. . One of the petals. . The stamens united into a monadelphous tube, which is cut open to show its form, the petals having been detached. . A longitudinal section of the calyx and ovary, showing two of its four cells, and surmounted by the style and disk, from the outer margin of which the staminal tube has fallen. А transverse section of the ovary, showing its 4 cells. . The fruit crowned by the persistent lobes of the calyx ; a quarter of the pericarp is cut away, to show the single suspended seed, marked by a groove left by pressure of the chord formed by the 3 abortive cells. . The fruit seen from below, showing its acutely quadrate form. . À transverse section of the mesopodal embryo, showing the line of junction of the exorhiza and neorhiza: all nat. size. —— ———— . The flower in bud of Agasta asiatica, with the bract at the base of the pedicel. . One of the four petals. . The monadelphous stamens combined at the base into a tube, which has fallen away from its attachment upon the outer margin of the disk. . The ovary, surmounted by the persistent lobes of the calyx, disk, and style. . À longitudinal section of the same, showing two of its four cells, the style, the disk with its inner margin expanded upwards, and its outer margin, from which the staminiferous tube has fallen. . А transverse section of the ovary, showing its 4 cells. . A fruit acutely 4-angled, the angles cordate at base upon the pedicel, surmounted by the persistent lobes of the calyx and the style: all nat. size. РтАтЕ XIII. A flower in bud of Butonica alba. 2. The same after expansion. . The same, seen from below, showing its 4-lobed calyx, its 4 petals with retroflected margins. The staminal tube with the petals attached, falling away from fig. 5. . The pedicel, ovary, and persistent calyx of the same, the corolla and stamens having fallen off. . А longitudinal section of the ovary, with the persistent asina lobes, disk, and style. . À transverse section of the same, showing its 4 cells. . А diagram showing the zestivation of the petals. А drupe of the same, according to the dimensions given by Rumphius (it appears smaller in his diminished plate 115) : a quarter of the pericarp is cut away longitudinally, to show the single suspended seed. À transverse section of the embryo, showing the line of junction of the exorhiza and neorhiza: all nat. size. A portion of the raceme of Butonica racemosa, with the flower expanded. А flower in bud, of the same. One of the four petals separated (which are attached by their claws to the staminiferous tube). : The stamens united at base into a monadelphous tube, which is cut open, and which has fallen off from the external margin of the disk. i Fig. 2 m 15. 16. 17, ‚ А full-grown fruit, from a specimen in the Museum of the Linnean Society. - A longitudinal section of the same, showing the form of its 4-winged pericarp, MR. J. MIERS ON THE BARRINGTONIACE Ж. 115 . The ovary surmounted by the calycine lobes and style, the staminiferous tube having fallen off. . А longitudinal section of the ovary, disk, and style, showing two of its four cells. . The drupe according to the shape and size given by Rheede, Roxburgh, and Hermann ; a quarter of its pericarp is eut away, to show the single enclosed seed : all nat size. . А portion of the raceme of Butonica Rumphiana with the flowers in bud. . À flower in bud on its pedicel. . The petals and stamens of the same, expanded, and agglutinated to the monadelphous tube, which has fallen from the external margin of the disk. . A longitudinal section of the ovary surmounted by the calycine lobes, disk, and style. . А transverse section of the 4-celled ovary. | ‚ The drupe, according to ће dimensions given by Rumphius, surmounted by the persistent caly- cine lobes; a quarter of its pericarp is cut away, to show the single enclosed seed. . A longitudinal section of the embryo, showing the neorhiza enclosed within the exorhiza : a// nal. size. PLATE XIV. . А portion'of the raceme of Butonica rubra, with flowers in bud, and another expanded. . À drupe of the same, according to the shape given in the diminished drawing of Rheede, but enlarged to its proper dimensions. . А transverse section of the same, showing within the pericarp the neorhiza of the embryo sur- rounded by its exorhiza: all nat. size. . A flower in bud of Butonica terrestris, with its calyx entirely closed. . The same fully expanded, as seen from below, showing the 3 ruptured lobes of the calyx, the 4 petals, and the stamens. . The stamens united into a monadelphous tube, which is cut open, showing two of the four petals agglutinated to it by their claws. . A longitudinal section of the 4-celled ovary, with two of the cells and their suspended ovules, two of the lobes of the calyx, the epigynous disk, from the outer margin of which the staminal tube has fallen off, the inner margin being raised, with the style seated in the hollow vertex. - À transverse section of the same. : 1 : . А drupe of the same, according to the dimensions given by Rumphius (it appears amaller in the diminished drawing of his plate 115) : a// nat. size. - А young bud of Butonica alata, with the calyx quite entire. - . The same somewhat grown, with the calyx beginning to split into 4 lobes. | : th i . A longitudinal section of the same, showing two of its four cells, the epigynous disk with its inner margin elevated. its single seed enclosing the embryo. The embryo extracted, with its exorhiza surrounding the neorhiza, and easily separable from it : all nat. size. i © The ovary, disk, calycine lobes, and style of Butonica rotata. ; : : b А fruit of the same, shown in longitudinal section, according to the dimensions given by : Sonnerat. 18. A transverse section of the same, including the seed : ай nat. size. — 116 MR. J. MIERS ON THE BARRINGTONIACEÆ. 19. The fruit of Butonica inclyta, according to the dimensions given by Griffith under the name of Barringtonia racemosa (but not of Blume) : this is is seen in his plate 636, $2, with analytical details, figs. 1-6, and is here seen in longitudinal section: the embryo of the seed is shown with its neorhiza beginning to separate from the exorhiza, the latter splitting into 4 bs at- both extremities, as in Dozomma cochinchinense. 20. A flower-bud of Butonica samoensis, the entire globular limb of its calyx being quite closed. 21. The ovary, with the limb of the calyx split into 3 lobes. 22. Тһе stamens combined into а monadelphous tube, to which one of the four petals is shown at- tached by its claw, all having fallen from the disk. 23. A longitudinal section of the ovary, with two of its three calycine lobes, the epigynous disk, on the outer margin of which the staminal tube was seated, its inner margin being considerably raised, forming the deep hollow in the vertex in which the style originates. 24. A transverse section of the ovary, showing its 4 cells. 35. The drupe, of the size and shape figured by Gaudichaud under Barringtouia racemosa: all nat. size. PLATE XV. i Fig. 1. A portion of the raceme of Megadendron macrocarpum, from Horsfield’s specimens and the excellent description of Hasskahl, showing flowers in bud, and another fully expanded, seen from beneath, with its 4 sepals, 4 petals with reflected margins, and expanded stamens. 2. Half of the staminal tube, viewed from within, to show the inner rows of capillary sterile short filaments, the more exterior series being longer and fertile. 3. A longitudinal section of the ovary surmounted by the calycine lobes, disk, and style. 4. A transverse section of the 4-celled ovary. 5. A drupe, of the shape and size described by Hasskahl, a quarter of the pericarp cut away, to show its single suspended seed. 6. The embryo of the seed, with its exhorizal portion beginning to split into 4 sections, as occurs also іп Loureiro's specimen of Дохопипа cochinchinense. 7. A longitudinal section of the same, showing the inner neorhiza ready to germinate. · 8. А transverse section of the same. 9. A leaf of Dozommo pendulum. 10. Portion of a raceme of the same, copied from a drawing of a living specimen, by Parish, with flowers in bud, and another fully expanded, seen from beneath, to show its 4 petals with reflected margins. ll. A fruit of the same, from drawings of Parish and Griffith in his plate 634 a, under Careya pendula : seen in longitudinal section, with its single suspended seed. 12. The seed detached. Fig. 1. 2. ‚ The embryo of the same, the testa being removed. . The embryo beginning to germinate. . The same germinating, seen in longitudinal section, showing the inner germinating portion (neorhiza) surrounded by the sphtting exorhiza, according to the analysis of Griffith: all nat. size. Prats XVI. A leaf of Dozomma cochinchinense е трна acutangula, Lour. ) , from Loureiro' s original specimen in the British Museum. А тш of the same, from Loureiro's specimen, with four undulated аш: ль с 11. 13. 13. 14. 15. MR. J. MIERS ON THE BARRINGTONIACE.E. 117 . А transverse section of the same, showing the embryo within the pericarp. . The embryo extracted, with the exhorizal portion splitting into 4 sections. . The same, seen in transverse section: all nat. size. . A leaf of Careya orbiculata from Griflith's specimen. . А flower, in bud, of the same. . The ovary of the same, from which the petals and stamens have fallen. . The ovary of Careya spherica, crowned by its 4 sepals and style. . The monadalphous stamens, with the tube cut open, to show that the more external and more internal series of filaments are capillary and without anthers, the intermediate series alone being fertile: one of the four petals remains attached by its claw to the tube. А longitudinal section of the ovary, showing two of its four cells, two of the sepals, the epigynous disk, on the outer margin of which the staminal tube was seated, the inner margin being raised, leaving a hollow in the vertex round the base of the style. А transverse section of the same. А drupe, crowned by the persistent sepals and style, a quarter of its pericarp being cut away to show the seeds imbedded in pulp, the latter having been removed. One of the seeds. The same, after germination, copied from Dr. Thomson's drawing: the neorhiza has sprouted into a rootlet at one extremity, while at the other it has thrown out a long naked stemlet, crowned at its apex by 4 young decussating leaves: all nat. size. PLATE XVII. Stravadium acutangulum, DC. Figs. 1 & 2. Drawing taken from the typical specimens of a Ceylon plant in the herbarium of Hermann, 15. upon which Linnæus established his Eugenia acutangula and Gaertner his Barringtonia acutangula, with which numerous other species have been confounded. . A flower of the same, in bud. ‚ The same, full-grown and expanded. The ovary crowned by the calyx: all nat. size. The same : magnified. . А longitudinal section of the same. і . The staminal tube cut open, showing the stamens and two of its four petals agglutinated to it by their claws, all having fallen off from their attachment to the disk. . А transverse section of the ovary, showing it to be only 2-celled : ай equally magnified. . The fruit, from Gaertner’s drawing. . The same, with half of the pericarp removed, to show the en повес send. . The embryo of the seed. 13. 1% A transverse section of the same, showing the exhorhizal portion splitting into 4 sections: А longitudinal section of the same, showing the neorhiza enclosed in the exorhiza. all nat. size. А portion of the plant of Chydenanthus excelsus, with its branching panicle, upon which an expanded flower is seen. 16. A flower in bud. 118 If 18. 19. 20. Fig. 1. 18. MR. J. MIERS ON THE BARRINGTONIACE Л. Тһе cup-shaped calyx, with a truncated margin : all nat. size. The monadelphous tube of the stamens, cut open, to show the inner row of short sterile filaments, the whole having fallen from the inner margin of the disk. А longitudinal section of the ovary, crowned by the calyx, disk, and style. A transverse section of the same, showing it to have only 2 cells: the last three figures are some- what magnified. Puate XVIII. A portion of the flowering specimen of Planchonia crenata, from Mr. Robert Brown’s typical specimen in the British Museum; the branching panicle shows a flower in bud and another fully expanded. . The stamens united into a monadelphous tube, upon which one of the four petals remains, to show how they are agglutinated to it by their claws, the tube having fallen from its attach- ment to the outer margin of the disk. . The ovary, surmounted by the sepals and style. . A longitudinal section of the same, showing two of its four cells, and surmounted by the stili; disk, and style. . A transverse section of the same, showing its 4 cells. . The drupe, drawn from Mr. Brown's specimen. A longitudinal section of the same, crowned by the persistent sepals, disk, and style, and showing | two of the four clls, with the seeds as they exist in situ, the pulp being removed. . А longitudinal section of another fruit, collected by Schomburgk ; this is gibbous, owing to the abortion of three of its cells, and it contains 4 seeds, horizontally attached to the axis, and apparently once imbedded in pulp, as in the preceding specimen. . One of the seeds, on its short thick funicle. . The same, deprived of its membranaceous integument and funicle, showing a depressed hollow on one side below the apex: seen on its face. . The same, viewed on its edge. . A longitudinal section of the same, seen from the interior of one of the faces, from which the loose embryo has been removed : it is an envelope of thick firm texture, opaquely white, seeming like an albumen. . The embryo extracted: all nat. size. . The same, magnified, showing the thick subtruncated terete radicle, upturned suddenly near the base, and supporting the two roundish fleshy subplicated cotyledons. . An edge view of the same: equally magnified. . А portion of a plant of Petersia africana, from a specimen given to me by Dr. Welwitsch. . А flower on Из pedicel. The ovary and calyx, the petals and stamens having been removed: all nat. size. 19 is an enlarged view of fig. 17, showing its 4-winged inferior ovary, crowned by the 4 eL init alternate 30. 21. Ап anther, seen before and behind : more magnified. 93. 25. 26. with the wings and the 4 petals. A longitudinal section of the ovary, disk, and style. А transverse section of fig. 20. The monadelphous stamens detached from the outer margin of the disk: all magnified. - A fruit, with its 4 large membranaceous wings. The pericarp of the same, deprived of its wings. A buxo section of the same: all nat. size. TRANS. LINN S0C.2. SER. Вот Tab. 10. EO ` M. ў / j / /, ffs Ж H / 7, /2 Ji 2 Maclure & Macdonald Lith, London . — y | J Miers, del. | BARRINGTONIA SPECIOSA. V aee ni S сн pn Масірге & Macdonald, lath. Lond on. АК АСТА GDIENLDHIDA Miers, del Die oaoa с R. 2 Bot. VOLL T45.12 STA ASIATICA. AGA pM — conta А. ~ 4 ASTA INDI G A mtt n е-е. лы ене" Масіпге k Macdonald, Lith. London: J.Miers del. D Blair Анод à SER. Вот. TAB 3 a TRANS. Linn. Soc оны YOR eS foe Maclure & Macdonald. Lath. London. J Miers, del. D Blair anto lith. Trans. Linn. Soc. 2. SER. Вот. ТАВ. М. С. Wira Па aded ie sv. TAB. | | 1 DA: TE T Вот. VoL 4 о ЕК. È TRANS. Linn Soc. SE Мала & Mardonala. lath london. TRANS. LINN. 506. SER. 2. Вот. VoL | TAB. [6. A PRIMORES CSS оф ауда ыты; са лылы жалы. а ага ЧЕК ғ. г ра, “ғ Ғе ТҮ Y " А D Ке: д Madue &MacdemiMbLeh Lemdon. о STPAVADIHM ACUTANGULUM. = “Меге del. D Выше auto Ju. TRans LINN. Soc. SER. 2. Вот. Vo, I. Tap 18 ОЕ J Miers del ТАРДЫ LANCHONIA CRENATA PETERSIA AFRICANA. Месте k Macdonald, Lith. London. E ІНЕ TRANSACTIONS OF THE LINNEAN SOCIETY OF LONDON. SECOND SERIES.—BOTANY. VOLUME I. PARI THE THIRD, MISSOURI BOTANICAL, GARDEN. BUN YO NR: PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. M.DCCC.LXXVI. VIII. IX. CONTENTS. PART III.—JvuwxE, 1876. On a Collection of Fungi made by Mr. SuuPIz Kurz, Curator of the Botanic Garden, Calcutta. Ву FREDERICK Currey, M.A., F.R.S., Sec. L.S. page 119 ‚ Preliminary Note on the rate of Growth of the Female Flower-stalk of Vallisneria spiralis, Linn. By ALFRED W. Веххетт, M.A., B.Sc., F.L.S. Lecturer on Botany at Si Thomas s Bopita > o . уу... Ж - On the Growth of the Flower-stalk of the Hyacinth. Ву Aurrep W. BENNETT, M.A., B.Sc., F.L S., Lecturer on Botany at St. Thomas's Hospital. . . 189 New British Lichens. By the Rev. W. A. Іжівнтох, В.А. Camb., F.L.S., етае Оп the Hygroscopic Mechanism бу which certain Seeds are enabled to bury them- selves in the Ground. Ву Еваксів Darwin, M.B., FLS. . . . . . 149 [ 19 | У. On a Collection of Fungi made by Mr. SuLrIz KUnz, Curator of the Botanic Garden, Calcutta. Ву FREDERICK Currtf, MA. Е.В.5., Sec.L.S. (Plates 19-21.) Read June 4th, 1874. THE fungi to which this paper relates were collected mostly in Pegu by Mr. Sulpiz Kurz, Curator of the Botanic Garden at Calcutta. І am not aware that any collections of fungi have heretofore been made in Pegu, and an interest therefore attaches to the present one as being so far the first of its kind. As might be expected, many well- known tropical forms occur. There are several species, especially amongst the Polyporei, which, although previously described, are by no means well known; and it will be seen that there are some new forms of considerable interest. In the examination of the plants I have had the advantage of access to Mr. Berkeley’s valuable herbarium, for which, and for his kind assistance in the determination of some of the species, I am desirous of expressing my acknowledgments. AGARICINI. Several species of Agaricus accompanied the other fungi; but it is hopeless to attempt to name specimens of this genus unless they are 2ccompanied by coloured drawings of the fresh plant, careful sectional figures, and notes of the colour and shape of the spores. Micrometric measurements of the spores should be added in all cases where practicable. Amongst the specimens were some which appeared to belong to 47. sil- vaticus, Fr., and Ag. fimicola, Bolt., and others which were identical with or allied to Ag. spadiceus, Fr., and Ag. picreus, Fr. Others, belonging to the subgenera Lepiota, Tricholoma, Volvaria, Psathyra, and Psathyrella, were not further determinable. Опе small Agaric is noted as being used for food. It resembles 4g. semiorbicularis, Bull. ; but that plant is not edible. Two species of Marasmius occur, one of which has not dried well: the other (No. 3472), in the dried state, is just like М. ramealis, Fr. ; but Mr. Kurz describes it as white when fresh, so that it can hardly be that species. Lentinvs cAPRONATUS, Fr. Epicrisis, p. 390; Persoon in Freycinet’s Voyage, t. 2. 1. 3.— Myodwine, Evergreen Forests, Pegu. No. 2604. : | This is ап interesting series of specimens, varying in the size of the pileus from $ inch to lj inch. In three of the plants the pileus is distinctly zoned. The gills cannot be called denticulate, although some are very slightly uneven at the edge. The figure in ‘ Freycinet’s Voyage’ is very good, but the specimens there represented are small. L. nEscENDENs, Fr. Epier. p. 390.— Tonkyeghat, Респ, Evergreen Forests. No. 2629. This species was found by Afzelius in Guinea, іп gravelly soil. It is figured by him in SECOND SERIES.— BOTANY, VOL. I. p 120 MR. Е. CURREY ON INDIAN FUNGI. the * Fungi Guineenses,’ t. x. fig. 22. Тһе specimen is in bad condition, and I have some doubts about the species. LENTINUS FURFUROSUS, Fr. Epicr. р. 391.—Yomah, Pegu. No. 3474 *. L. vELUTINUS, Fr. Epier. p. 392.—Yomah, Pegu. No. 3498. L. GLABRATUS, Mont. Cuba, p. 424. No. 1779.—North Rajmehal hills, N. Bengal. There are but two specimens of this species, and they have not been very successfully dried. Тһе margin of the pileus can hardly be called hairy ; but in all other respects the plant agrees with Z. glabratus. L. Sazon Салт, Fr. Еріст. p. 899. Rumph. Amb. t. 56. f. 1.—South Andaman. Nos. 2560, 2561. “ Sajor Саји” is the Malay name of the plant. L. ехилз, Fr. Epicr. p. 393.— Tonkyeghat, Nakawa, Evergreen Forests, Респ. No. 2592. . Sittang valley, Pegu. Nos. 3487 © 3488. L. rNQUINANS, Berk., Hooker’s Journal of Botany, 3rd series, vol. iv. p. 130, t. vii. f. 1.— Kemendine, Jack Tree Gardens, near Rangoon, Pegu. №. 2600. à I think there is no doubt about the species, although the single specimen gathered by Mr. Kurz varies from the type in some particulars. Тһе gills are somewhat narrower, and the warts and the brown pubescence of the pileus have rubbed off in drying; but there are dark spots all over the pileus, showing the places where the warts were attached. The bristles have disappeared, except here and there at а short distance from the margin, Where they are still plainly visible under a lens, and the transverse scales on the stem are not now to be seen. Тһе latter, however, like the warts, may have been rubbed off in drying. Mr. Berkeley has noticed that in some dried specimens a pellicle separates from the umbilicus, carrying the warts with it and leaving the pileus smooth. L. KunzrANUS, n. sp. Pileo lento, infundibuliformi, furfuraceo-squamuloso, fusco; stipite brevi, fusco-ferrugineo, ad basin nigricante ; (те profunde decurrentibus, leete rubiginosis.—Yomah range, Респ. Хо. 3497. This is a beautiful species, in many points resembling L. furfurosus, Fr., from which it differs in the deeply funnel-shaped pileus and in the colour of its gills. In ZL. furfurosus the mycelium (held together by the sand in which it grows) forms an indurated tuberi- form root. In the specimens of the present species the stipes has the appearance of having been broken off in gathering; so that it is not improbable that a tuberoid radix may be present in the perfect state, which would bring the plant still nearer to Z. fur- furosus, Fr. The colour of the gills varies from a bright rust-colour to a greyish brown, according to the incidence of the light under which they are examined. Plate 20. fig. 11. Plant, natural size. L. casprrosus, n. sp. Dense сюзрНозиз. Pileo infundibuliformi, squamis subconcen- tricis ornato; stipite versus basin attenuato; lamellis substipatis, decurrentibus, (siccis) brunneis.—Pegu. No. 2624. No special locality is given. * In the specimens of this species the stems have been broken off short in gathering, so that there is no tuberiform radix, and the gills are too dark-coloured to be described as “ flavidis ;” but the plant comes so near in other respects to L. furfurosus, Fr., that I do not venture to propose it as a new species. 4 МВ. Е. CURREY ON INDIAN FUNGI. | 1M I find no species described which agrees with this. The plant is stated to be white when fresh. Іп the dried state it is of a brownish fawn-colour, the gills being somewhat darker than the pileus. Plate 19. figs. 4 and 5. Fig. 4, a group of plants, natural size; fig. 5, a single plant. LENTINUS IRREGULARIS, n. sp. Pileo (sicco) nigricante, fragili, infundibuliformi ; Jamellis cinnamomeis, irregularibus, hie per totum stipitem decurrentibus, illie annulatim abruptis.—Karen country. Pine-forests from Theemeechu to Borkee. Хо. 2534. There are four specimens of this species, in two of which the gills run down the whole length of the stem, whilst in the other two they terminate abruptly, and break off ina way very similar to what is seen in the British species Г. fimbriatus, Curr. If the two latter specimens only had been examined, I should probably have referred them to L. fünbriatus. Plate 19. figs. 14 and 15. Plants, natural size. SCHIZOPHYLLUM COMMUNE, Fr. Several specimens of this species occur in Mr. Kurz's collection.—The localities are:—Botanie Garden, Calcutta, Nos. 1781 and 2649; Ross Island, South Andaman, No. 2562; Sikkim Himalaya, Nos. 2572 and 2575; Tonkyeghat, Pegu, No. 2598. LzNzrrES ALBIDA, Fr. Epicr. р. 405.— Tonkyeghat, Pegu. Хо. 2601. I have seen no specimens of L. albida, Fr. Хо. 2601 agrees with the description in the * Epicrisis, except in being stipitate and in the gills having а cinnamon-colour. Nothing is said by Fries as to the size of 7. albida. The pileus in the present plant measures 2 of an inch by 2 of an inch, and at first sight much resembles Panus stypticus, Fr.; but the anastomosing gills show that it eannot be that species. L. Ратлвотт, Fr. Epicr. p. 404.—Seven Pagodas, Evergreen Forests, Tonkyeghat, Реби. Хо. 2590. | There is another specimen in the collection (No. 2604), which seems to be the same species, from the Botanic Garden, Caleutta. POLYPOREI, Fr. PoLvronvs (Мвворов) PERENNIS, Fr. Epicr. 484; ib. Hym. Eur. 531. Sent by Dr. Stoliczka from Penang. Хо. 2041. P. (MzsoPus) XANTHOPUS, Fr. Epicr. 437.—Pegu, Nakawa, Evergreen Forests, Tonkye- ghat, No. 2533; Sittang valley, Pegu, No. 3489; Yomah range, Pegu, No. 2627. There is also a specimen from South Andaman, No. 2563. P. (Мкворгв) FLORIDEUS, Berk., Hooker's Journal of Botany, 3rd series, vol. vi. р. 187.— Pegu, Karen country, Bookee ridges, 5000—6000 feet, Pine-forests, No. 2595. This species was found by Dr. Hooker at Darjeeling, at a height of 7500 feet. P. (MzsoPus) нүроргАвтиз, Berk., Hooker's Journal of Botany, vol. viii. p. 174.—Howrah district. On bamboo. Хо. 2652. , This species was first found by Mr. Spruce at Panuré. Mr. Berkeley has noticed that the pileus is sometimes lateral; and this is the case with three out of the four specimens sent by Mr. Kurz. Judging from Mr. Kurz's specimens alone I should have placed the Plant in the division Pleuropus. 4 5 122 MR. Е. CURREY ON INDIAN FUNGI. POLYPORUS (MEsoPUS) CRASSIPES, n. sp. Pileolignoso, infundibuliformi, fusco-purpureo, margine umbrino et zonato; stipite durissimo, badio, ad basin incrassato; poris minutissimis, stipite pallidioribus.—Y omah river, Реса. There is no number to the specimen. This plant resembles in some points P. parvulus and Р. carbonarius, Fr., but differs widely in other respects. I do not know any species to which it can be referred. Plate 19. fig. 13. Plants of the natural size. P. (PLEUROPUS) SANGUINEUS, Fr. This common tropical species occurs in Mr. Kurz's collection from Mutlah, in Lower Bengal, No. 2031; from the Botanic Garden, Calcutta, No. 2682; and from Tongloo, Sikkim Himalaya, 5000—6000 feet, No. 2584. P. (PrgvROPUS) морезѕтоѕ, Kunz, Fr. Linn. v. р. 519; Fr. Epicr. p. 444.— Yomah range, Northern Pegu, No. 2610. This is the same species as Р cervino-nitens, Schweinitz. P. (PLEUROPUS) FLABELLIFORMIS, Kl., Fries, Epicr. p.444.—Y mah range, Pegu, No. 3491. P. (PLEUROPUS) AFFINIS, Fr. Epier. p. 445.—Yomah range, Респ, Nos. 2628 & 3490. Evergreen Forests, Bookee ridges, Karen country, 5000—6000 feet, No. 2617. No. 2617 has a very short stem and an irregular margin, but I cannot make it out to be any other species. P. (PLEUROPUS) РТСТРЕЗ, Fr. Epicr. p. 440.—Botanic Garden, Calcutta, No. 2655. There is also a specimen from Sikkim without a number. P. (PugvRoPUs) LUCIDUS, Fr. Epicr. p. 442.—Botanie Garden, Calcutta, No. 2684. Тһе yellow form of the species. P. (PLEUROPUS) AMBOINENSIS, Fr. Epicr. p. 442.— Karen country, Респ, 4000 feet. Хо. 2636. This is a very variable species, probably not distinct from P. lucidus, Fr. Р. (PLEUROPUS) SPLITBERGERI, Mont. Cent. ii. no. 83; Sylloge, p. 164. no. 522.—Nat- toung Hills, Pegu, 6000-7000 feet. Хо. 2638. Plate 19. figs. 2and 3. Fig. 2, the under side of the plant ; fig. 8, the upper side: natural size. This Polyporus agrees almost to the letter with the description of P. Splitbergeri in the * Sylloge, although the character * contextu aureo-nitido " is hardly applicable to the present plant. In the figure the pileus appears zonate, but the apparent zones are rather imbrications produced by the successive growths of the pileus. In Polyporus Nilgheriensis, Mont. (Sylloge, p. 164. no. 523), а closely allied species, the pileus is de- scribed аз * antice prole nova subimbricata," and this description is exactly applicable to the present specimen of P. Splitbergeri. This plant being stemless, it might be thought that it ought to be placed in the division Apus; but it clearly belongs to the “ proli- ficans ” division of Fries, as described in his * Novee Symbolæ Mycologice,’ p. 161, № which division the stem is sometimes wanting. : P. (PLEUROPDS) BRUNNEO-PICTUS, Berk., var.—Arracan, Kolodyne valley, on old wood. No. 2045. Plate 19. figs. 11 and 12. Fig. 11, the under side ; fig. 12, the upper side: natural size. MR. Е. СОВВЕУ ON INDIAN FUNGI. 123 This plant comes so near to P. brunneo-pictus, Bk., that І do not venture to propose it as а new species. It differs in its obsolete stem and in the nature of its pores. In P. brunneo-pictus the apices of the pores are papular, with a central aperture; and slight traces of this structure may be seen at the extreme margin in Mr. Kurz's plant, in which, however, the pores are not generally porotheloid, although the dissepiments are unusually thick. P. brunneo-pictus is described in Hooker's Journal of Botany, 3rd series, vol. viii. p. 176. PornvPoRvs (APUS, ANODERMEI) FUNALIS, Fr. Epicr. 459.—Lower Bengal, Mutlah tidal jungles. No. 2029. This species is admirably figured in Afzelius’s * Fungi Guineenses,’ t. i. f. 3. Р. (Аргв, ANoDERMEI) RUBIDUS, Berk. London Journ. Bot. vol. vi. p. 500,—Pegu. Nos. 2621, 2622. | Plate 20. figs. 5 and 6. Fig. 5, the upper side; fig. 6, the under side of a plant: natural size. The hymenium differs little in colour from the upper side of the pileus. Тһе pores are very small, just visible to the naked eye. "The plant varies in the brightness of its tint, although always exhibiting a tinge of flesh-colour. In one of Mr. Kurz's specimens the pilei are imbricated, and one has a distinct lateral stipes, showing a variability of habit. The species was first found by Gardner at Point de Galle, Ceylon, on fallen trees. P. (ANODERMET) INCERTUS, n. sp. Pileo in centro tenuissimo, marginem versus іпегаѕ- sato et reflexo, leviter zonato, dilute cervino; poris amplis, profundis, irregularibus, cervinis. Pileus very thin, almost obsolete, except at the margin, where it is thicker, reflexed, slightly zoned, and .fawn-coloured. Pores large, deep, irregular, somewhat darker than the pileus. Plate 19. figs. 6 and 7. Fig. 6, a portion (rather more than half) of a plant, natural size; fig. 7, a section. This is a difficult plant to describe technically on account of its great irregularity of growth. It has the appearance of having grown in an almost resupinate position, but with the margin to a certain extent free all round. The figure will give a better idea of it than any amount of description. Its systematic position is doubtful. It is certainly not one of the veritable ** Resupinati;" but whether it belongs to the ** Anodermei " or the “ Inodermei ” is very difficult to decide in the dry state. It has no appearance of having ever been “ succose ;" but, on the other hand, I see no trace of a cuticle. Р. (PLACODERMEI) APPLANATUS, Fr.—Pegu, Tonkyeghat, Evergreen Forests, Tonnghoo. No. 2686. P. (PLACODERMET) MARGINATUS, Fr.—Bookee, Pine-forests, 4000-6000 feet, Karen country, Pegu. No. 2586. P. (PLacopERMEI) Ревзоомп, Fr.—Pegu, Nakawa, Evergreen Forests, Tonkyeghat. No. 2598. P. (Рглсоревмет) нотозстивоз, Berk. Lond. Journ. Bot. vol. vi. p. 479.—Pegu, Yomah, Myodwine, Evergreen Forests. Хо. 2614. 124 MR. Е. CURREY ON INDIAN FUNGI. PornyroRUs (APUs, PLACODERMEI) CINEREO-FUSCUS, n. sp. .Pileo dimidiato, lignoso, durissimo, margine tenui, basin versus incrassato, fusco-cinereo, rugoso-tuberculoso ; poris minimis, cinereo-fuscis. Pileus 11-2 inches wide, of а dull brown colour, with blotches of black, thin at the margin but thickening rapidly towards the base, longitudinally rugose with tubercles; pores very minute, blackish brown, sometimes almost quite black.—Pegu, Nakawa, Ever- green Forests, Tonkyeghat. Хо. 2606. | Mr. Berkeley was inclined to refer this plant to his Polyporus ferreus (London Journ. Bot. vol. vi. р. 502. по. 175); but it differs from that species іп its dark pores and in some other characters. It is allied to P. fasciatus and Р. supinus of Swartz. Plate 19. fig. 1. Plant, natural size. Р. (INODERMEI) CINNABARINUS, Fr. Epicr. p. 473.—On old logs, Botanic. Garden, Cal- cutta. No. 2681. Р. (INoDERMEI) scRUPOSUS, Fr. Epicr. p. 473.—Martaban Hills, Shan toung gyre toung, 3800 feet. No. 3492. P. (INopERMEI) Fret, Fr. Epicr. p. 476.—Pegu, Yomah. No. 3493. Р. (INoDERMEI) VERSICOLOR, Fr. Ерісг. p. 478.—Sikkim Himalaya, 7000 feet, on dead wood. No. 2567. P. (INopERMEI) PINSITUS, Fr. Epicr. p. 479.—Nattoung ridges, Karen country, 6000 feet. No. 2608. А. (INODERMEI) CINERASCENS, Fr. Ерісг. p. 481.—Yomah, Northern Pegu. Хо. 2609. Fries describes the pileus in his P. cinerascens as fuscous, and the pores as cinereous. In Mr. Kurz's plant the pileus is rather ochraceous than fuscous, and the pores are reddish brown. These differences in colour are hardly sufficient to separate it as a species. P. (INODERMEI) ANEBUS, Berk. Lond. Journ. Bot. vol. vi. p. 504.—Tonkyeghat, Pellowa, Evergreen Forests, Pegu. Хо. 2599. This is a small thin form of Mr. Berkeley's species. From Mr. Berkeley's remarks (loc. cit.) it seems to be a variable plant. P. (INoDERMEI) XEROPHYLLACEUS, Berk., Hook. Journ. Bot. 3rd series, vol. viii. p. 200. №. 2658.—Botanic Garden, Caleutta, on old logs in the rainy season. Mr. Kurz's specimens agree very well with Mr. Berkeley’s description. In some the pilei are confluent laterally, in others imbrieated. І should rather have placed the plant in the “ Anodermei," but have retained Mr. Berkeley's division, Plate 20. figs. 1 and 2. Plants, natural size. TRAMETES LOBATUS, Berk., Hook. Journ. Bot. 3rd series, vol. iii. p. 84.—N. Bengal, Sikkim Terai, on old logs. Хо. 2569. _ T. UMBRINUS, n. sp. РИео resupinato, effuso-reflexo, lobato, subtomentoso, leviter zonato ; poris amplis, dentato-laceris. Хо. 2611.—Pegu, Nakawa, Evergreen Forests, Ton- kyeghat. Plate 21. fig. 4. Plants, natural size. It is difficult to say whether the larger specimen is a single plant or a mass of MR. Е. CURREY ON INDIAN FUNGI. 125 confluent pilei. The pores underneath the slightly reflexed marginal lobes are at right angles with the pileus. In the centre the pores become elongated and decurrent, over- lapping one another in an imbricated manner. TRAMETES OCCIDENTALIS, Fr. Epicr. —Botanie Garden, Caleutta. Оп old logs in the rainy season. Хо. 2656. I have had a little doubt about this species, as the pileus is not “ effuso-reflexed," which is one of the characters of T. occidentalis. In Mr. Kurz's specimens the pilei are dimidiate. Substituting ** dimidiato-" for “ effuso-reflexo " in Fries's description, the latter applies exactly to Mr. Kurz's plant. T. orNGULATUS, Berk., Hook. Journ. Bot. 3rd series, vol. vi. p. 164.—Pegu, Yomah. No. 3475. Other specimens in Mr. Kurz's collection from the Botanic Garden, Calcutta (Nos. 2679 & 2685), show that this species varies a good deal in the thickness and іп the pubescence of the pileus, the latter being sometimes silky, sometimes woolly in its appear- ance. Тһе pubescence, of whichever kind, is always slight, a lens being almost neces- sary to see it. distinctly. DADALEA. DupALEA TENUIS, Berk.—Pegu, Yomah. Мо. 2608. Another specimen of this species from South Andaman Island occurs in Mr. Kurz's collection. D. ртѕсогов, Fries, Ерісг. p. 494.—Pegu, Yomah. No. 3508. D. 7охата, Fr: Epier. p. 494.— Pellowa, Evergreen Forests, Tonkyeghat, Респ. No. 2589. I have not seen specimens of Fries's D. zonata. The present plant only differs from Fries's description in the pileus being wrinkled and very slightly pubescent, characters which hardly justify its being made a new species. I should place it under D. zonata as уат. rugoso-pubescens. HEXAGONIA, Fr. НЕХАСОМТА РОТУСВАММА, Mont. Fr. Еріст. p. 497; Mont. Sylloge, p. 169; Cuba Crypt. (Ramon de 1а Sagra), р. 379, t. 14. f. 3.—Pellowa, Evergreen Forests, Tonkyeghat, Pegu. No. 2597. H. TENUIS, Fr. Epicr. p. 498.— Респ, Elephant Point. On old trees along the shore. No. 3484. This is exaetly the plant of Fries, agreeing in every particular. There are two other forms (both from Yomah) in the collection (Nos. 3494 & 3505), which I cannot satis- factorily separate from И. tenuis. No. 3494 differs in having a darker-coloured pileus and smaller pores, the latter being brown, without the grey tinge of the pores of И. tenuis. In No. 3505 the pileus and pores are precisely similar in colour to those of No. 3494, but are а good deal larger. 126 MR. Е. CURREY ON INDIAN FUNGI. НЕхАвОхтА KURZI, n. sp. Pileo coriaceo, glabro, orbiculato aut reniformi aut dimidiato, corrugato, zonato, rubro-fusco; alveolis amplis, intus canescenti-fuscis, acie cinna- momea.—Mutlah. No. 2028. Plate 20. figs. 3, 4, and 7. Figs. З and 7, the upper side; fig. 4, the under side of different plants : natural size. This species comes very near to И. nitida, D.R. et M. It differs in having the pileus sometimes quite circular, and in the colour of the pores, which are of a pale bright grey in the inside, with the edge of a distinctly cinnamon-colour. HYDNEI | Нурхом сом, Fr. Ерісг. р. 517—Mutlah, Calcutta. No. 2030. TREMELLODON GELATINOSUM, Fr. Hymenomycetes Europæi, p. 618.—Sikkim, 7000-8000 feet. No. 2578. Tremellodon is a genus formed by Fries to include Hydnum gelatinosum, Scop., and Н. auriculatum, Fr. These plants have the aculeate hymenium of Hydnum, but the fructification of Tremella. See Linn. Journ. vol. v. p. 181. ТЕРЕХ FLAVUS, Kl.; Fries, Epicr. p. 522.— A rracan, Baronga Islands, оп old wood, 1000 feet. Хо. 2047. I. PALLESCENS, Fr. Epicr. p. 522.—Yomah, Pegu. №. 3495. There is one other Zrper in the collection (No. 3486) from the Sittang valley, Pegu. It has been partially destroyed by insects, and is not certainly determinable; but it must, I think, be either T. flavus, or Г. sinuosus, Fr. GRANDINIA GRANULOSA, Fr. Epicr. p. 527.— Pellowa, Evergreen Forests, Tonkyeghat, Респ. Ко. 2594. AURICULARINI. THELEPHORA PUSILLA, n. sp. Pusilla, fuscescens, simplex vel ramosa, sursum cristato- divisa.—Sikkim Himalaya, 2000-2500 feet. Хо. 2570. This species is evidently allied to 7. cristatella, B. & Br. (a Ceylon plant), but it differs in the colour, which varies from light to dark brown, and in the divisions, which are rather clavate than acute. Plate 21. fig. 9. Plants, natural size. T. PALMATA, Fr. Epicr, p. 537.—Botanic Garden, Calcutta. On bamboo-trunks. No. 2661. STEREUM ELEGANS, Fr. Epicr. p. 543.—In bamboo-jungles, but the locality is not given. Хо. 1792; Lower Pegu, junction of Pegu river and Khayasoo Choung. No. 2681. The specimens under the latter number are very small, but there is no doubt about the species. S. OSTREA, Fr. Epicr. 547.—Sikkim Himalaya, 4000-6000 feet. Мо. 2579. 8. LoBATUM, Fr. Epicr. p. 547.—Pegu, Tonkyeghat, the seven Pagodas. Хо. 2596. In Mr. Kurz's collection there is another specimen of 8. lobatum (No. 2564), from Escape Bay, South Andaman Island. МБ. Е. CURREY ON INDIAN FUNGI. 127 STEREUM SCYTALE, Berk., Hook. Journ. Bot. 3rd series, vol. vi. p. 170.—Pegu, Karen Hills, Bookee ridges, 5000-6000 feet. Хо. 2618. Респ, Yomah. Хо. 3504. Plate 19. figs. 8, 9, and 10. Еіс. 8, the under side; figs. 9 and 10, the upper side of different plants : natural size. This species was found by Drs. Hooker and Thomson in the Khasia mountains, and by Capt. (now General) Strachey in the Western Himalayas. S. ADUSTUM, Lev. Ann. d. Se. Nat. (1844), 3rd series, vol. ii. р. 218.—Southern Yomah, Pegu. No. 2602. This is more properly a Hymenochete. 5. PAPYRINUM, Mont. Sylloge, p. 178.—Pegu, Timeokee, Evergreen Forests. No. 2623. S. MEDICUM, n. sp. Effusum, resupinatum, tenuissimum, margine subhirsuto; hymenio inequali іп prominentiis rotundis elevato, fusco-tabacino, subvelutino.—Sikkim, 5000-6000 feet. Хо. 2582. SEX This is а very curious production: the pileus is very thin, but the plants grow together in a stratified mass, the mass having exactly the appearance of being composed of thin layers of tobacco. Itis used by the Lepchas in medicine, and has a native name. Plate 20. figs. 9 and 10. Fig. 9, a portion (about half) of a plant ; fig. 10, a section : natural size. S. CYATHIFORME, n.sp. Cyathiforme, extrinsecus hirsutum, margine inflexo; hymenio levi ad marginem cyathi pallide umbrino, basin versus rubro-fusco.—Pegu, Karen Hills, Bookee ridges, 5000—6000 feet. Хо. 2619. Plate 21. fig. 1. Plants, natural size. This is a very pretty plant, looking likea Crucibulum without sporangia. CoRTICIUM VIOLACEO-LIVIDUM, Fr. Epicr. p. 564.—Pegu, Yomah. No. 3476. С. Laviearum, Fr. Epicr. p.565.—Pegu, Yomah. Хо. 3507. C. Могвкотп, Fr. Epicr. p. 558.—Sikkim Himalaya, Phalloot, on dead trees, 11,000- 12,000 feet. No. 2571. GUEPINIA RAMOSA, n. sp. Czspitosa, simplex vel ramosa; ramis apice bifurcatis vel cristatis; hymenio aurantiaco ; stipite subtus cinerascente.—Arracan, Kolodyne valley, on old wood. Хо. 2044. Howrah district, on old logs. No. 2640. Plate 21. figs. 2and 3. Plants, natural size. Varying from 1 to 1 inch in height, and from the simple forms of fig. 3 to the ramose state of fig. 2 CLAVARIEI. ÜALOCERA viscosa, Fr. Epicr. p. 581.—Seebpore, Lower Bengal. Хо. 2196. Txrnvra ruscrezs, Fr. Epicr. p. 586. À specimen of this species occurs with No. 2567 (Polyporus versicolor). TREMELLINI. TREMELLA FOLIACEA, P. Fr. Syst. Myc. vol. ii. p. 212. —Sikkim Himalaya, on dec 10,000 feet. Хо. 2580. SECOND SERIES.—BOTANY, VOL. I. = 128 MR. Е. CURREY ON INDIAN FUNGI. HINNEOLA AURICULA JuD#, Fr. Hym. Eur. p. 695.—Nakawa, Evergreen Forests, Ton- kyeghat. No. 2591. Н. AURICULA Canis, Fr. Syst. Мус. vol. ii. p. 222 (sub Hvidia)—Pegu, Northern Yomah, on logs. No. 2615. NIDULARIACETI. CYATHUS INTERMEDIUS, Mont. Sylloge, р. 284.—Seebpore, Lower Bengal. No. 2198. This plant agrees very well with Montagne’s species, but it is not striate within. I do not observe the blackness in the substance of the sporangia noticed by Montagne ; but this feature probably depends upon age. LYCOPERDACEI. LYcOPERDON GEMMATUM, Fr. Syst. Myc. vol. iii. p. 36.—Sikkim Himalaya, 7000-8000 feet. Хо. 2576. Spores globose, nearly 0:0002 inch. No. 2581, also from Sikkim Himalaya, 9000— 10,000 feet, and No. 2653, from the Botanie Garden, Caleutta (in bamboo-thickets amongst dead bamboo remains), may be forms of L. gemmatum, Fr.; but there are no notes, and dried Zycoperdons are most difficult to determine. L. PUSILLUM, Fr. Syst. Myc. vol. iii. p. 33.—Lower Респ, Kayosoo Choung, in elephant- grass jungles. Хо. 2630. L. BRASILIENSE, Fr. Syst. Мус. vol. iii. p. 40.—Pegu, Nakawa, Evergreen Forests, Ton- kyeghat. №. 2635. The specimens are, unfortunately, in very imperfect condition; but the peridium is membranous and punctate, the flocei broad and lax, and it grows on wood. It seems to . agree sufficiently well with Fries's species. Itis accompanied by Hypoxylon marginatum, Sehw. | TRICHODERMACET. TRICHOCOMA PARADOXUM, Jungh. Sikkim Himalaya, 6000-7000 feet. On mossy dead stems. No. 2585. MYXOGASTRES. TRICHIA PYRIFORMIS, Hoffm. On wood, with No. 2585 (Trichocoma paradoxum). ARCYRIA UMBRINA, Fr.—Pellowa, Evergreen Forests, Tonkyeghat, Pegu. Хо. 2626. HELVELLACEI. PEZIZA AURANTIA, Pers. Fr. S. M. ii. p. 49.—Sikkim, great road, on earth. No. 2578. P. RUTILANS, Fr. S. М. ii. p. 68.—Junction of the Choungmenal Choung and Tonkyeghat rivers. Оп mud banks. Хо. 2587. HELOTIUM crrrinum, Fr. Peziza citrina, Fr. S. M. ii. p. 131.—8ikkim Himalaya, 5000- 7000 feet. On rotten wood. №. 9583. МБ. Е. CURREY ON INDIAN FUNGI. 129 SPOROCYBACEI. STILBUM INCONSPICUUM, n. sp. Solitarium vel cæspitosum, minutissimum ; capitulo subgloboso, albido ; stipite badio, deorsum leviter incrassato ; sporis ellipticis, nucleo excentrico ornatis, 0:0002 unc. longis.—Seebpore, Lower Bengal. No. 2197. S. ERYTHOCEPHALUM, Dittm. var .—On dead bamboo-stems, Botanic Garden, Calcutta, rainy season. No. 2639. This plant hardly differs from S. erythrocephalum, Dittm.; but the stem is only very delicately pilose, with silky adpressed hairs. SPHJERIACEI. XYLARIA DIGITATA, Grev. Spheria digitata, Fr. S. M. ii. p. 326.—Botanic Garden, Cal- cutta. Оп moist logs. Мо. 1791. X. TABACINA, Kickx.—Sikkim Himalaya, 7000-8000 feet. On old fallen trees. No. 2575. X. GUYANENSIS, Mont. Sylloge, p. 202.—Tonkyeghat, Evergreen Forests, Pegu. No. 2625. X. roLYMORPHA, Grev. Spheria polymorpha, Fr. S. M. ii. p. 326.—Botanie Garden, Caleutta. On oldlogs in the shade, rainy season. Хо. 2659. X. HYPOXYLON, Grev. Spheria hypoxylon, Fr. S. M. ii. p. 327.—Arracan, Baronga Island, on old tree-stumps. Хо. 2046. Botanic Garden, Calcutta, on old logs, rainy season. No. 2657. X. KunzrANA, n. sp. Solitaria vel cæspitosa, vel etiam connata, 1-1 unc. longa; capite eastaneo subhemispherico, peritheciis prominentibus scabro; sporidiis nigro-fuscis, amygdaloideis, 0:0006-0:0007 unc. longis.—On brick-laid paths where fire had been burnt. Botanic Garden, Calcutta, rainy season. No. 2650. Plate 21. fig. 8. Plants, natural size. seda The dried plants required to be moistened to bring out the colour, which is a chestnut- brown. Mr. Kurz describes them as yellow at first and then brown. There is another Aylaria in the collection (No. 1783, from the Calcutta Botanic Garden), which seems to be the same as No. 2650. The habitat, viz. brick-laid soil upon which fires had been burnt, is somewhat peculiar for a Xylaria. X. FLAGELLIFORWIS, n. sp. Solitaria vel cæspitosa, simplex vel fureata, flagelliformis, colore carneo ; peritheciis ?—0On earth, Sittang valley, along the base of the Yomah range, Респ. No. 3485. Plate 21. fig. 5. Plants, natural size. i X. MUTABILIS, n. sp. Solitaria vel cæspitosa vel connata, simplex vel ramosa, vel cristata ; stipite nigro, rugoso, interdum profunde sepulto; capite elongato, үне carneo, sicco pallide umbrino, ostiolis nigris prominentibus scabro ; sporidiis uniseriatis, nigro-fuscis, irregularibus (reniformibus, subellipticis vel subglobosis), minutissimis, vix 00001 unciæ attinentibus.—On brick-laid paths in the Botanic Garden, Cal- eutta. Very common in the rainy season. №. 1790. Plate 21. figs. 10 and 11. Plants in different states, natural size. This plant is remarkable for its irregular growth, the small size of its sporidia, and the T2 180 MR. Е. CURREY ON INDIAN FUNGI. densely crowded state of the asci, which are often difficult to distinguish. I do not know any species of Xylaria with similar fructification. HYPOXYLON SUBORBICULARE, Welw. & Curr. Trans. Linn. Soc. vol. xxvi. p. 281.—On old logs, Botanie Garden, Caleutta. Хо. 1809. Mr. Kurz's specimens exhibit in section a browner and more eorky substanee than the typical H. suborbiculare; but I see no reason to doubt the identity of the plants. Тһе difference is probably due to age. Н. concentricum, Bolt. On old wood, Botanic Garden, Calcutta. No. 1810. Н. MARGINATUM, Schw. This species occurs (as above mentioned) upon the wood to which Lycoperdon brasiliense was attached. DIATRYPE RUGOSA, п. sp. Irregularis; stromate rugoso vel depresso et marginato, intra et extra nigro; peritheciis in stromate sepultis, plus minus biserialibus; sporidiis irregularibus, globosis, ellipticis vel pyriformibus, 0:0005—0:0006 unc. longis. On hard wood. Yomah, Реза. No. 3471. Тһе stroma in this species occurs in minute rounded patches, sometimes it is pro- minent and rugose, sometimes depressed, with a margin, and then (under a lens) looking very like a Patellaria. HYPOCREA VARIABILIS, n. sp. Statu juniore placentiformis, disco vel umbone concen- trico ornata, colore citrino; statu adulto subglobosa vel irregularis, rugosa, colore umbrino; sporidiis Нар. On living leaves of bamboo, Yomah, Pegu. Мо. 3480. Plate 21. figs. бапа 7. Fig. 6, part of a leaf with plants, natural size; fig. 7, а plant, magnified. Э5РН ЖЕТА PHASELINA, Mont. Sylloge, p. 239. Hab. Nakawa, Evergreen Forests, Tonkyeghat, Pegu. S. sUBLIMBATA, D. В. et Mont. Fl. Alg. p. 498; Sylloge, p. 228. Apparently on stems of Thysanolena acarifera, Kambala toung, top. MIcROPELTIS APPLANATA, Mont. Sylloge, p. 245. On Girroniera leaves, Mangrove Bay, South Andaman Island. Хо. 2565. FUMAGO SALICINA, Tul. Sel. Fung. Carp. vol. ii. p. 280. On Sphenodesma erycibordes, Pz., Wa-choung, Yomah, Pegu. No. 3482. Хкствта EUGENLE, п. sp. Peritheciis pallide citrinis, tomentosis, іп stromate concolori collocatis vel ejusdem marginem amplectentibus ; sporidiis Р On dead leaves of Eugenia. Yomah, Респ. М№о. 3479. This is a very pretty species. The perithecia for the most part are seated round the base of the stroma, the colour of which is rather darker than that of the perithecia. GnAPHIOLA РнкЕхісів, Poit.—On leaves of Phanix paludosa. Calcutta. No. 1744. The systematic position of this curious plant has not yet been satisfactorily determined. Some years ago l examined with great care some specimens kindly furnished to me by the late Sir William Hooker, and the results of my observations were published in the Microscopical Journal in 1859 (1st series, vol. vii. p. 295). Му conclusions, however, do not coincide with those of Montagne, whose aecount of the plant is to be found in the * Annales des Sciences Naturelles, 4th series, vol. xii. p. 188. E — Ч ЖУ AY?) " ШУ E шж ҮҮ e шалы” MR. Е. CURREY ON INDIAN FUNGI. 131 DESCRIPTION OF THE PLATES. АП the figures except fig. 7, Plate 21, are drawn to the natural size. PLATE 19. Fig. 1. Polyporus cinereo-fuscus, n. sp. Figs. 2 and 3. Polyporus Splitbergeri, Mont. Fig. 2, the under side ; fig. 3, the upper side. Figs. 4 and 5. Lentinus cespitosus, n. sp. Fig. 4, a group of plants ; fig. 5, a single plant. Figs. 6 and 7. Polyporus incertus, n. sp. Fig.6, a portion (rather more than half) of a plant; fig. 7, а section. Figs. 8, 9, and 10. Stereum scytale, Berk. Fig. 8, the under side; figs. 9 and 10, the upper side of different plants. Figs. 11 and 12. Polyporus brunneo-pictus, Berk., var. Fig.11, the under side; fig. 12, the upper side. Fig. 18. Polyporus crassipes, n. sp. Figs. 14 and 15. Lentinus irregularis, n. sp. PLATE 20. Figs. 1, 2, and 8. Polyporus xerophyllaceus, Berk. Figs. 8, 4, and 7. Hexagonia Kurzi,n. sp. Figs. З and 7, the upper side; fig. 4, the under side of different plants. Figs. 5 and 6. Polyporus rubidus, Berk. Fig. 5, upper side; fig. 6, under side. Figs. 9 and 10. Stereum medicum, n. sp. Fig. 9, a portion (about half) of a plant; fig. 10, a section. Fig. 11. Lentinus Kurzianus, n. sp. PLATE 21. Fig. 1. Stereum cyathiforme, n. sp. Figs. 2 and 3. Guepinia ramosa, n. sp. Fig. 4. Trametes umbrinus, n. sp. Fig. 5. Xylaria flagelliformis, n. sp. Plants, natural size. 4. Figs. 6 апа 7. Hypocrea variabilis, n. sp. Fig. 6, part of a leaf with plants, natural size ; fig. 7, a plant magnified. Fig. 8. Xylaria Kurziana, n. sp. Fig. 9. Thelephora pusilla, n. sp. Figs. 10 and 11. Xylaria mutabilis, n. sp. Plants in different states. In some of the specimens, notably in the large specimen on the right of fig. 11, fragments of brick remain attached to the bases of the plants. [.18.] VI. Preliminary Note on the rate of Growth of the Female Flower-stalk of Vallisneria spiralis, Linn. Ву AvrnED W. BENNETT, M.A., B.Sc., F..L.S., Lecturer on Botany at St. Thomas’s Hospital. Read November 4th, 1875. ALTHOUGH the extraordinary rapidity of growth of the peduncle of the female flower of Vallisneria spiralis (probably one of the most remarkable instances to be found in the vegetable kingdom) appears to be familiar to botanists, I have been unable to find any record of actual measurements. Тһе present communication to the Society I prefer to call a “ Preliminary Note,” inasmuch as a much more extensive series of experiments than I have at present had the opportunity of making will be required before the subject can be considered exhausted. The object of the great length attained by the female flower-stalk is well known. The flower is by this means brought to the surface of the water in which it grows, which I should infer, from the length it attains (though I do not recollect any exact record of this), to be from three to four feet. It is there fertilized by the pollen supplied by the male flower, which, growing submerged on a short peduncle, breaks off from it and rises to the surface, where it floats about until the pollen meets with a female flower, the latter after impregnation being again carried, by the spiral coiling of the peduncle, beneath the surface, where it ripens its seeds. The first memoir, as far as I have been able to trace, of the structure of Vallisneria is by Е. J. Quekett, ** Observations connected with the Anatomy and Physiology of Vallis- neria spiralis," in the first and only volume of the * London Physiological Journal’ for 1843. Тһе mode of fertilization and the phenomena of rotation within the cells of the leaves are here described; and the description appears to be the basis of the ordinary accounts found in works on descriptive botany. А far more minute and accurate account of the structure of the plant is by Chatin, in his ‘ Mémoire sur le Vallisneria spiralis, Linn., considéré dans son organographie, sa végétation, son organogénie, son anatomie, sa tératologie, et sa physiologie, 1855. "The author of this memoir points out the singular error into which previous writers have fallen, in describing the peduncle of the female flower as originally coiled up spirally and unrolling towards the period of impregnation, an error repeated by Richard, Turpin, and even by Grenier and Godron in their * Flore de France, and by nearly all subsequent writers, whether scientific or popular, and forming the text of one of Erasmus Darwin's most роейса] images in his “ Loves of the Flowers. Chatin states, and with perfect correctness, that “la hampe des fleurs femelles, d'abord droite, ne se déroule jamais." Neither of these authors, however, gives any details on the special subject of this paper. The first flower noticed on the plants growing in my aquarium, where the water is about 8 inches deep, was on July 19th of the present year, when the total length of the 184 MR. А. W. BENNETT ОМ THE RATE OF GROWTH OF THE peduncle was 26 in., lying in several long coils on the surface. I am almost certain that the whole of this growth, at all events in excess of the 8 in. which would have brought the flower-bud to the surface, had taken place within the preceding 48 hours. At 10 o’clock the next morning the peduncle was again measured, and had attained the extraordinary length of 38 in., or an increase of 12 in. in 24 hours. At 4 р.м. it measured 41 in. ; at 9 А.м. on the 21st 421; and at 10 А.М on the 22nd it had attained its ultimate length of 43 inches. Тһе window in which the aquarium was placed had a south-west aspect; but the climatal conditions were by no means favourable to rapid growth, the temperature being rather low, the sun scarcely visible during the whole time, and the rainfall excessive. Lest it should be thought that this is the result of an | error of observation, I may mention that Mr. W. W. Reeves, the Secretary of the Micro- scopical Society, informs me in a note that he has observed an even greater rapidity of growth than this. I do not know whether any other instance is known of a single internode attaining a length of over 34 feet. This flower was never fertilized (there being at that time no male plant in my aquarium), but remained open, and the stigmas in an apparently receptive condition, nine days longer, till July 31st, when I left home. During this time no actual recoiling of the peduncle had taken place, though it dis- played a strongly marked waviness. It did not occur to me, until the flower-stalk had nearly completed its growth, that it would be interesting, in accordance with the plan proposed by Prof. Sachs, to ascertain in which portion of the flower-stalk the main portion of this rapid increase took place, though a few measurements which I did make convinced me that the apical portion was that in which there was the greatest activity. І was unable to renew my observations before the latter part of September, in which interval a large number of female flowers had appeared. Тһе one on which the most complete series of observations was made was first observed on Sept. 21st, at 10 A.M., when the apex of the flower-bud was 5 in. below the surface of the water, or the flower- bud and peduncle measured together 3 in. In two days and а half, or by 4 Р.М. on the 24th, the peduncle had reached the length of 10-7 inches, or, when straightened, the base of the unopened flower-bud was 2:7 inches above the water, the bud, enveloped in its spathe, being 0:5 in. in length. Тһе following were the successive measurements :— inches, Increase. Peduncle and bud.......... Sept. 21st, 10 a.m. 3 й 22nd, 10 a.m. 4:75 175 „ 12 пооп 5 “25 » 2 P.M. 52 : » 8 P.M. 5:5 3 ^" 10 р.м. 5°75 "25 » 23rd, 9 a.m. 7°25 1:5 » 12 noon 7.6 35 ” З P.M. 8:0 4 Peduncle only ........ ....Бері. 23rd, 6 pm. 8-0 5 - 10 р.м. 8-4 4 ” 24th, 8.30 л.м, 10-0 1-6 ” 4 р.м. 10-7 7 FEMALE FLOWER-STALK OF VALLISNERIA SPIRALIS. 155 During the three portions of days constituting this period, the growth was at the rate respectively of 1°75, 2:5, and 3:25 in. in twenty-four hours. At this stage it seemed possible to measure the rate of growth of different zones of that portion of the flower-stalk which, when straightened, was exposed above the surface. As, however, neither of the **auxanometers," or instruments for measuring growth, described by Sachs in his ‘ Text-book’ can be used for weak stems which do not natu- rally stand erect, the only available mode appeared to be to mark off at intervals of time measured lengths from the base of the flower-bud. This was done (but necessarily, from my engagements, at very unequal intervals) with black varnish, each length of one inch being marked off as it rose (when straightened) above the surface of the water, and indicated successively as А, B, C, &c., any fraction ofan inch remaining over being added to the subaqueous portion. Тһе following table exhibits the result, neither the final length attained nor the rapidity of growth being so great as in July, as, indeed, might have been expected. n B UU B B FF M Н 4 КЕ L NM Бош ТИН. DE Sept. 24th, 4 r.m. 1:0 1:0. 87 107 ДАУА 20: 19 ^ T9 80 119 8 Ж Ө Жи ЕСІ ЕЕ 19 .. Di OL OM мат ш) 81 1 314 27,1043. 225 2:5. 1-8 1:0 10 10. 0.10: 10 10.10, ... 88.2995 1096 2,00xX. 20 27519 L15 115 11 105 POB F0 10.190 10 BL. 2485. 15 Юри. 27520 20 115 PIGS EFI 14 ri ыы r05 55 2015 25 oma, Fiw 20 80 21 PIS 115 PIS 146 11 ТҮ Та 11 11 87b Жуда 4 10 р.м. 31 81 215 1:15 115 115 1:15 1:15 1°15 1:15 1:15 1:15 875 2745 `6 20th, 9 л.м. 8-25 8:93 2-15 1-15 1-15 1:15 115 1:16 1:15 1:15 1:15 115 88 2775 3 10 рм. 3:25 3:25 2-15 1-15 1:15 1:15 115 1:15 115 115 1:15 115 88 2775 0 The general results arrived at are that between 4 р.м. оп Sept. 24th and 9 А.М. on Sept. 29th the flower-stalk had increased in length from 107 to 27°75 in., the two zones immediately beneath the flower-bud from 2 in. to 6'5 in. This latter portion had in- creased during this period 225 per cent., the remainder of the flower-stalk 144 per cent. of its original length. | The term “ Energy of Growth " of any particular part or zone of a plant is defined by Sachs (* Text-book,' p. 741) as *the power of that partieular zone to attain a definite length." In the case of the stem of a plant consisting of a number of internodes, the measurements of Sachs, Pfeffer, and others show that the greatest energy of growth is generally displayed by an internode situated at some distance from the terminal bud, the energy of growth decreasing in the successive internodes, both towards the apex and towards the base of the stem. Similar careful measurements do not appear to have been made of the relative energy of growth of different portions of the same internode. In the case of the greatly elongated internode of Vallisneria, the maximum energy would appear to be very near the apex, the total energy of growth of the two terminal zones being to that of the remaining portion as 225 to 144, or nearly as 3 to 2. The energy would appear also, as far as any conclusion can be drawn from my few measurements, to decrease towards the base. For, dividing the portion of the flower-stalk excluding the SECOND ЗЕВТЕ$. —ВОТАМУ, VOL. I. U 136 MR. A. W. BENNETT ON THE RATE OF GROWTH OF THE two terminal zones again into two parts, the uppermost of these (C—L) increased be- tween 10 a.m. on Sept. 27th and 9 a.m. оп Sept. 29th, from 9:8 to 11:35 in., or 15:8; the lowermost of them from 8:8 to 9°95 in., or only 13 per cent of its original length. In ealeulating what Sachs terms the ** Grand Curve of Growth " of the plant *, I have pursued a somewhat different plan to that adopted by this high authority. In his ‘Text-book’ (p. 737) he gives the following instructions for the construction of such a eurve:—** If the increments of growth during successive equal times are drawn as ordinates, with the intervals of time as abscissze, a curve will be obtained which, starting from the axis of abscissæ, reaches a maximum of elevation and returns again to the axis.” In constructing these curves, it appeared to me that results of greater value would be obtained if, instead of the aetual inerement for any given time, the propor- tionate increase in relation to the length at the commencement of the interval were taken as the ordinate. Ву this means it seems to me that the energy of growth of each particular portion of the stem at any particular time will be represented by the altitude of the curve. Thus, supposing a stem increases during two successive equal intervals of time from 4 to 5 and from 5 to 6 inches, the energy of growth will not be equal to the two intervals, but will be less in the latter interval in the proportion of 4to 5; inasmuch as it will have increased during the former interval 25 per cent., during the latter interval only 20 per cent. of its length at the commencement of that interval. It will be evident that though my grand eurves of growth will agree with Prof. Sachs's in their general features, they will differ in minor particulars, and will in general be less abrupt. In constructing the curves from the data already given, I have first of all divided the increment during each interval of time by the length at the commence- ment of the interval, in order to get the proportionate increase as above indicated, and then the result by the number of hours in the interval, in order to reduce the times to a uniformity, the length of the intervals between the observations varying greatly, owing to my engagements. It is obviously, then, indifferent on what scale the curve is constructed, 4. е. what is taken as the unit of measurement. In the diagram which accompanies this (р. 137), each unit of 1 inch on the ordinates represents an increase of 1 per cent. per hour; while an inch on the line of abscissee represents a period of 12 hours. In accordance with Prof. Sachs's plan I have placed each ordinate in the centre of the abscissa to which it belongs. The two accompanying curves have thus been obtained, one exhibiting the variation of the energy of growth for the two terminal portions, the other for the remainder of the flower-stalk. Тһе general result is seen to be that the terminal portion exhibits a considerably greater energy during the whole of the time with the exception of the first few hours. А second series of observations was made on another flower-stalk; but I have not thought it worth while to record the results in detail, in consequence of an accidental injury (to which the slender and delicate stems are extremely liable when undergoing measurement), which both retarded the growth and caused it to be somewhat irregular. The eurves construeted in the same manner showed a striking resemblance to those in the previous instances, but with two abrupt depressions caused by the injury referred to. * See * Arbeiten des botanischen Instituts in Würzburg, 1873, Heft 2, t. i. 157 ' . А ce à T" 3 « P. s ба Curves of Growth of the two terminal zones and the remaining portion of the female flower-stalk of Vallisneria spiralis, Linn, Б А. Grand eurve of growth of the two terminal zones; B. Grand eurve of growth of the remaining portion of flower-stalk. Оп the line of abscisse every 2 ineh represents a period of 12 hours; on the ordinates every inch represents a growth at the rate of 1 per cent. per hour. The figures above = each ordinate express the growth of curve A, those below the ordinate the growth of curve B on this notation. The ordinates are erected from the ш centres of the several periods of time, a, b с, &c. c» Е < сі i [22] $ ' Z (xm : 5 ane і — bud i "i AU ; > тч р > - ©! 29 ' “; e Тез ау, ' nd m Фе, : << = р са E А , ' p = 1 . . ec Н : : = , E: : ки = ' { 2 i ; < m ' ES Е iS Ries i N T 24 x is oe m i ы : р у hse ы 8 io = : T ELI - 7" 2 ao нт а стали 4 рэа eno i 5 à » ; т = ° т ж = = = қ < 25 - = т ж - Put < (> ES Y © оо © Q 2 e = Я = d ci 5 3 5 5 " t É = Е й d E ái 138 ON THE FEMALE FLOWER-STALK OF VALLISNERIA SPIRALIS. This flower-bud was first observed at 10 А.М. оп Sept. 22nd, when its apex was 1 inch below the surface. Ву 9 А.м. on the 23rd, the total length of the peduncle was 9:75 ір. ; at 8.30 a.m. on the 24th, 13:95 in. ; 8.30 a.m. on the 25th, 16:6 in.; 10 А.М. on the 27th, 19:0 in. ; and at 9 А.М. on the 28th it had reached its ultimate length of 19:75 in. Two terminal zones, together 1:6 in. іп length, were first marked off at 6 Р.М. on Sept. 23rd, and had finally, at 9 a.m. оп Sept. 28th, attained the length of 3:6 inches, or had inereased 125 per cent. on the original length. Тһе remainder of the stem had during the same time increased from 10 in. to 16:15 in. or 61:5 per cent. The total energy of growth of the terminal portion was therefore in this instance just double that of the remainder; but the latter had undoubtedly been checked by the accidental injury. Тһе flower-bud, when it fivst rose to the surface of the water in the aquarium, was ‘Sin. in length, and closely enveloped in its spathe, from which it shortly afterwards gradually emerged, but did not open until the growth of the flower-stalk was nearly com- pleted, when it had attained a length of "75 in. Тһе flower remains open floating on the surface for weeks, with the stigmas in an apparently receptive condition. Ву the time its growth is completed the flower-stalk has become very wavy, but does not coil up in shallow water when the flowers are unimpregnated. Тһе only record I have been able to find of observations on the rate of growth of different portions of the same internode is by J. Münter, in the first volume of the * Botanische Zeitung’ for 1843. Не here lays down the general law that in the inter- nodes of dicotyledonous plants the energy of growth, or, аз he terms it, the vis procreativa, increases from below upwards, or may be described as centrifugal. This result he found most strikingly displayed in the peduncle and pedicels of Pelargonium, the only observations, I believe, which have hitherto been made on different parts of indi- vidual internodes *. It is interesting to compare this with the law of growth of stems taken as a whole, which is thus laid down by Sachs (* Text-book,’ p. 740) :—“ As it is usual for several contiguous internodes of stems to be growing at the same time, and the maximum rapidity of growth occurs, according to circumstances, іп the 2nd, 3rd, 4th, or 5th internode beneath the bud, the region of most rapid growth is at a considerable distance from the apex of the stem, and especially when the internodes attain a con- siderable length and several are growing at the same time. Іп roots, оп the other hand, the maximum rapidity of growth occurs much nearer the punctum vegetationis, usually at a distance of only a few millimetres.” It would appear, therefore, as if the phenomena of growth in the flower-stalk of Vallisneria exhibit a closer similarity to those of the roots than of the aerial stems of terrestrial plants. * See also the reference to Miinter’s observations on the Hyacinth, postea, p. 143. [ 199 3 Pa ~~ УП. On the Growth of the Flower-stalk of the Hyacinth. By Атғкер W. BENNETT, M.A., B.Sc., F. L.S., Lecturer on Botany at St. Thomas's Hospital. Read March 16th, 1876. IN a paper which I had the honour of reading before the Society at its meeting on November 4th, 1875, I gave some details in respect of the remarkably rapid growth of the flower-stalk of the female flower of Vallisneria spiralis. The general results arrived at were that the greatest “ Energy of Growth” was displayed by the apical portion of the peduncle or that immediately beneath the flower-bud, the energy apparently decreasing regularly towards the base of the flower-stalk. As this appeared to be opposed to the law stated by Sachs and others to govern the rate of growth of the different successive internodes of an aerial stem, I was anxious to ascertain how far it was in aecord with the relative rapidity of growth of different portions of a single elongated aerial internode. Гог this purpose I have taken the earliest opportunity during the present spring of measuring the growth of the common peduncle of the inflorescence of the Hyacinth, with the following results in two specimens, one grown in a hyacinth-glass, the other in soil in a pot. Specimen А, grown in a hyacinth-glass. This was first measured at noon on Feb. 23rd, when the peduncle, with a total length of 1:25 in., was divided into two equal portions of 0:625 in. At 10 А.м. on the 26th, when it had increased to 1:55 in., each of the two sections was again divided, the length of the four portions, proceeding from above downwards, being 0:35, 0:4, 0-4, and 0-4 in. Measurements were made twice and sometimes three times a day; and it was soon evident that the energy of growth of these different portions was very unequal. By 10 р.м. on Feb. 29th each of the three uppermost portions was still only 0°5 in. long, whilst the lowest had increased to 1:0 in. From this time the increased rapidity of growth of the lowest portion was still more marked. By 10 r.m. on March 5th the lengths were respectively 0-9, 0-9, 0°85, and 2°35 in., and at 10 р.м. on March 11th, when the growth had finally ceased, the measurements were 1:15, 1:0, 1:0, and 3:45 in., making à total of 6:6 in. The following is a complete table of the measurements :— A B C D Total. Increase. Fob. Теке "35 4 4 4 1:55 19 e 0. "85 4 4 4 е 05 Юри. 85 4 4 5 1-65 :05 nh, Ти... 4 4 4 -55 1:75 1 о. 4 4 4 -55 1:75 | 28th, 10 a.m. . 45 4 + "65 1-9 15 Зи an 45 4 4 vfi 1:95 -05 140 ON THE GROWTH OF THE HYACINTH FLOWER-STALK. А В C D Total. Increase. КОЮ S. и... 45 45 -4 "75 2-05 1 А... -5 45 -45 8 22 15 SIX 2. -5 5 45 -95 2-4 :3 ШЕРИ Ны, 5 5 5 1:0 2:5 `1 March 186, ли ........ Dw D 59 5 1.2 2.8 3 11252 ee `6 55 5 1°25 9:9 1 Оне, IDEEN ues -65 6 -55 із 8-1 "92 о. у б 65 "55 1:35 3°25 15 IR РАҚ... ‘75 65 "6 1-5 3:5 95. SEMIS eens ies ‘75 7 $ ү 3:75 :35 ДО жи 521.22... 8 xi 6 1-9 40 25 ZR IAM. а... 8 75 “75 2.2 4-5 “5 о... “85 8 "8 28 4-75 25 БЫ АИ о. -85 8 "8 9:8 4-75 ; РИ у у t. 9 9 “85 2:35 5-0 :25 АЕ... ‘95 "9 9 2:5 5:25 "25 Bras оо. 1:0 9 "9 № 5:5 “25 ТШ. ЧО MONS ROME лы сол 1:0 9 "9 9-75 5-55 :05 ҮК... 1:0 9 95 2-9 5°75 2 ӨШ, Fe AMS оаа. 1:05 "9 1:0 8:0 6:0 :25 о Ве Г Т, ы. 1-1 1:0 1:0 3:15 6:25 “25 Oth; Тоха UT 11 1:0 1:0 9:3 6:3 :05 Hby3952.2i0 c5 1:15 1-0 1:0 8-25 6-4 1 TO MODO IM 1:15 1:0 1:0 8-4 6-55 15 А... D 1-0 1:0 8-45 6:6 “05 ЕСЕГЕ cU 1:15 1-0 1:0 3°45 6-5 ‘0 It will be seen from the above table that by far the greatest total energy of growth was displayed by the lowest of the four segments, which increased during the twelve days between Feb. 26th and March 10th from 0-4 to 3:45 in., or 762°5 per cent. of its original length. The next greatest energy, but at a great interval, was exhibited by the apical section, which increased from 0°35 to 1:15, or 228 per cent., while the two central portions exhibited the least activity, increasing only from 0:4 to 1:0, or 150 per cent. of their original length. Dividing the peduncle into an upper and a lower half, and constructing the “ Curve of Growth” in the same manner as those given for Vallisneria (p. 137), the accompanying diagram (Curve A) represents the results diagrammatically. On the ordinates, again, 1 in. represents a growth of 1 per cent. per hour on the length at the commencement of each interval; but each inch on the line of abscissee represents а period, not of 12 but of 24 hours. Іп order, therefore, to compare the curve of growth of the Hyacinth with that of Vallisneria, the former ought to be flattened out so as to occupy twice the space. It will be seen that the rate of growth was subject to great variations on different days, dependent, no doubt, on the temperature and other causes. Making the division between day and night at 10 a.m. and 10 р.М., the growth was nearly equally divided between them, the 5 inches growth between Feb. 26th and March 10th being accounted for by about 2:65 in the daytime and 2:35 at night. — Specimen А. The thicker curve represents the Energy of Growth of Sections C & D, the thinner i ! "7 ^ : : : | curve the energy of growth of Sections А & В; the figures above the ordinates < 2%; А represent the percentage growth per hour during the respective intervals-of the y о ! 3 ; Ñi former, those below the ordinates that of the latter curve, On the line of ! өсімдер деа 2 : | absciss® every inch represents а period of 24 hours, A.B { . * 1 . «о a с.р -- i^ ! 3 ЕБ $ E: \ glo Biz АВ. SiN с i BIS =18 ШЕ мін 518 : Š + + , + оте L \ —— + 1 > 4. d e © ES 9 © - ж e 2 = = д Е- ‚я ES = еп E Е e E Е E t E 4 $ E E 43 E = К = а = = ж = ж я = = z | р р | i . Specimen В. à The thicker eurve represents the Energy of Growth of Section D, the thinner curve the ! ^ * . | energy of growth of Sections А, В, & С; the figures above the ordinates repre- ! . . . ! sent the percentage growth per hour during the respective intervals of tho ! : hi former, those below the ordinates that of the latter curve. On the line of ] in! : ; | FQ ! ! abseisss every inch represents a period of 24 hours. ! 1 ! І р - | 1 I \ i Si I l °з - | і ! ( a e І У) ! р — Ң i | ] | 2 1 ©! ! | ] FHS ©! e р o ! І : S 519 : ә d © оо f ! ә? т | ) Жс ес 15 Ц ! Qe | р i + АВ.С. Е : > ' I {© 1 is г 21 2 i i а L 1 . А 1 № 1 А 1 1 15 3% Ж 15 "n = м 4 т T d o Lg T T т T E 1 са ea < wD © i 00 e = 3 S 2 с СЯ © d БЫ d 5 Е = ж = = 2 я = м д © 143 MR. А. W. BENNETT ОМ THE GROWTH OF THE Specimen В, grown in a pot. In the second example, the evidence was still more conclusive that the growth of the peduncle is mainly basilar. On Feb. 26th, the flower-stalk, then an inch in length, was divided into two equal portions of 0:5 in. On the next day, when it had increased to 1-1 in., the lowest zone of 0:1 in. was marked off separately. By 10 р.м. on Feb. 29th this lowest zone (C+D) had increased to 0-7 in., or by 600 per cent. of its original length, while the two uppermost zones were still respectively only 0:55 and 0:5 in. long. Тһе lowest zone was then again divided into two portions, the upper one being 0:5 and the lower 0:2 in. long. By 10 р.м. on March 3rd the lengths of the four zones, commencing from the top, were 0:8, 0:8, 075, and 0°75 in., giving а total of 3'1 іп. At 10 р.м. on the 7th, the total length of 6:5 in. was distributed thus :—1:6, 1:5, 1:25, and 2:15 in. ; and at 10 А.М. on the 13th, when the final length of 8:2 inches had been attained, the measurements were respectively 2:2, 1:75, 1:5, and 2°75 in. The following is the complete table :— i А В С D Total. Increase. Жоқ QUE. DAE ch a ow. 5 5 1:0 ОВИ ЯСИ Б "5 1:0 "0 ZUR IRE ea -5 -55 a 1:05 05 о. 5 5 d 11 05 ЖШ бе а LE 5 5 2 1.2 1 S33. 5 01s 5 5 “2 12 0 S eS 5 "5 9 1-2 0 ZA ва... "55 “5 "45 5% 15 2 Жа. -55 5 Б @ 1-75 25 Мана талаған. 55 55 "5 3 1-9 15 ве... 6 55 "55 3 2-0 1 Ed Жал. 22. 6 6 55 4 2-15 “15. EX uou 65 6 -55 5 2.3 15 Me A cee. s, "85 "65 “65 "6 9-55 25 в. S “75 “75 "65 "65 98 :25 и... "8 8 т 75 3-1 8 Zi Ол... Kk D .9 8 1-1 37 "6 ыла о. "9 :9 85 1:25 3:9 2 COMEDIT и 1:0 1:0 1:95 4-95 85 ие. 1-05 iu 1:1 1:25 45 25 Oth 1048 205.55 y 1:25 l1 1:55 5:1 6 Ири. у E 1:35 1:8 1-2 1-75 5.6 | Е САМ uui 1:5 1:5 1:25 1:85 6-1 5 Мру ОТ 1-6 1:5 1:25 2:15 6:5 4 Sth JO с 1:65 1:5 1:25 2:15 6:55 05 JUGE. а "S IT 1:55 1:35 2-15 6:75 2 SM te VL L.S 1:85 1:6 1:4 9:15 7:0 95 IDEM ns 2-0 1:65 1-4 2-35 7:4 4 И “2725 ОН 9-0 1:65 1:4 2:45 7:5 d IR 22-22 29 1:65 1:4 2:45 15 0 DARE ш... 2-0 1-65 1-45 2-55 7-65 15 IB IARE, ус. уг: л: 2-15 Pr 1:5 2-75 8-1 45 Е Gp 0 OD 2-15 I 1:5 2.15 8-1 0 ІБ, ТОЛИ о 22 1-75 1:5 2-75 8:2 I Шри oo 2.2 1-75 15 9-75 8-9 0 FLOWER-STALK ОЕ THE HYACINTH. 143 Starting from the measurement at 10 р.м. on Feb. 27th, the lowest of the three zones, which then measured 0:1 in., had increased by March 13th so as to make up the two zones C & D together 4:25 in., or 4150 per cent. of its original length, while the remainder had only increased from 1:0 to 3:95, or at the rate of 295 per cent. Again, starting from 10 р.м. on March 29th, when the four zones were first marked off, the ultimate increase of the lowest was from 0:2 to 2:75 in., ог 1275 per cent.; the next greatest energy was displayed by the uppermost, which increased from 0°55 to 9:2, or just 300 per cent. ; next came the second zone from the top, which showed an increase from 0:5 to 1-75, or 250 per cent. ; and finally the third from the top, showing an increase from 0°5 to 155 in., or exactly 200 per cent. The curve B in the accompanying diagram (p. 141) illustrates the total energy of growth of Section D as compared with that of Sec- tions A, B, and С taken together, the former being, as in the case of the previous curve, in excess during almost the whole period of growth. The rate of growth was again subjeet to great irregularities, which will be seen to correspond to a considerable extent to those of curve A, and were no doubt attributable mainly to changes in temperature. Making the division between day and night as before at 10 А.М. and 10 r.m., the total amount of growth was again not very different in the two; but instead of being, as in the previous case, slightly in favour of the day, was rather more decidedly in favour of the night; of the 6:5 in. growth from Feb. 29th to March 13th, 3 inches was by day, and 3:5 inches by night. It will therefore be seen that, as far as these observations on the relative growth of different portions of the same internode go, they are entirely in accord with the statement of Prof. Sachs (vide ante, p. 135), in regard to that of different internodes on the same branch, that the maximum energy of growth is exhibited at a period considerably below the punctum vegetationis, though it is here much nearer the base than in the cases measured by Sachs. "This brings out into still stronger relief the opposite phenomenon displayed by the elongated submerged flower-stalk of Vallisneria, the energy of growth of which is manifested mainly in the apical portion. Тһе elongation of the peduncle of the Hyacinth continues considerably after the complete expansion of the flowers, until the lowest in the raceme begin to fade. These observations differ in several points from those on the flower-stalk of the Hyacinth recorded by Münter in the * Botanische Zeitung ' for 1843, Feb. 24; but as he gives no measurements, I cannot think that these latter are of very great weight. Не deseribes its growth as not centrifugal, like that of most flower-stalks, but centripetal ; that is, it сеазез to grow first near the flower and finally at the base. It will be seen that шу two experiments (and I think the care with which the measurements were made precludes any possibility of mistake) agree in this, that while the energy of growth is Sreatest in the basal portion, the apical portion continues to grow for very nearly or quiteas long. Тһе growth of the flower-stalk of Pelargoniwm he describes, on the other hand, as centrifugal, the growth of each zone ceasing before the one next above it. With regard to the relative amount of growth by day and by night, Münter also gives по measurements, but states that in the daytime the plant grows at first five times, then four times, and then three times stronger than by night. This differs materially from SECOND SERIES.— BOTANY, VOL. I. E wsi ON THE GROWTH OF THE HYACINTH FLOWER-STALK. the general law as stated by Sachs (‘'Text-book,’ English ed. p. 749 et seq.), that “the plant will, according to circumstances, sometimes grow more quickly by day, sometimes by night, without exhibiting any exactly recurrent periodicity,” the difference, however, being never so great as that stated by Münter. My own observations are more in accordance with this. Since writing the above, my attention has been called by Prof. Sachs to a series of papers by Reinke in the * Botanische Zeitung’ for the present year, on the phenomena of growth of stems. His experiments were made entirely on the flower-stems of Endogens, chiefly Juncus, Scirpus, and Narcissus; but the only point that bears on the present inquiry is the statement that “ the part of Narcissus in which growth takes place is entirely beneath the earth, and of the rushes within the leaf-sheath, all the parts which rise above the earth having already completed their growth." Unfortunately Reinke does not give the measurements on which he founds this statement; and, if correct, it presents а singular want of harmony with the law of growth exhibited in the Hyacinth. The point seems to deserve further careful investigation. А very few measurements which I made on the leaves of the Hyacinth indicate that the increase in length in them takes place entirely in the basal portion, at least after they have attained a considerable length (see Sachs, *'lext-book,' English ed. p. 137). Whether, however, the cell-division is carried on actually beneath the surface of the soil, as stated by Reinke to be the case in the stem of Narcissus, I am unable to say. [ 145 ] = Pd L^ VILI. New British Lichens. Ву the Rev. W. A. LEIGHTON, В.А. Camb., F.L.S., F.B.S. Ed. Plate XXII.) Read February 3rd, 1876. VERRUCARIA MYRIOSPORA, Leight. Pale yellowish, filmy, very thin, effuse or deter- minate by a pale brown watery line ; apothecia shining black, moderate in size, nume- rous, round or oblong, often difformed by confluence, sessile and conspicuous, hemispherico-subpapillate ; epithecium a very minute pore in the apex of the papilla ; perithecium dimidiate, spreading at the base, carbonaceous ; asci narrow, lineari- clavate; paraphyses few, very slender and delicate; spores innumerable, colourless, very minute, oblong or lineari-oblong, simple. On young oaks, Kylemore and Doughruagh, co. Galway (October 1875), Mr. Lar- balestier. A singular and peculiar species, differing in its general aspect and its innumerable very minute spores from any described species. Spores “003 mm. long, "0015 mm. broad. Spermogonia numerous, minute, roundish or clavate, black like the apothecia. Sper- matia large, linear, arcuate. Gelatina hymenia I—. VERRUCARIA SUCCINA, Leight. Thallus fuscescent, thin, effuse, scarcely if at all sur- rounding the base of the apothecia; apothecia amber-colour, numerous, large, hemispherico-conical, papillate; perithecium | amber-colour throughout, dimidiate, spreading at the base; epithecium a minute роге in the apex of the papilla; para- physes very delicate and slender; asci lineari-clavate; spores 8, colourless, broadly Jusiform, 7-septate, large. On moist rocks near the lake, Kylemore, co. Galway, very rare, Mr. Larbalestier (1875). The apothecia, when wetted, became of a beautiful transparent amber-colour. Spores ‘046 mm. long, “008 mm. broad. No spermogonia detected. Iodine has very n if any reaction on the asci and spores. Its nearest ally would seern to be F. illinita, N уі but that species differs in having a whitish thallus, nigricant apothecia, the perithecium colourless in the lower part, and fuscous in the upper part, and iodine батын the asci and spores of a dirty deep brown, and is also а corticolar lichen. y. succina differs also in the size and colour of the apothecia from 7. chlorotica (Ach.), which is consociate on the same rock, and which has also smaller 3-septate fusiform spores. LECIDEA SUBDILUTA, Leight. Thallus of a dark dirty sea-green colour, thin, effuse, xen tinuous, very slightly rimulose; apothecia pale, waxy, minute, elevato-sessile ; epi- thecium pale brown, plane or concave, surrounded by а thickish, pale, waxy, elevated я х9 146 THE REV. W. А. LEIGHTON ON NEW BRITISH LICHENS. margin; hypotheciwm colourless, or only very slightly yellowish at the very base; paraphyses indistinct, conglutinate, apices brown; spores 8, colourless, oblong, 1-septate, somewhat constricted in the middle. On maritime rocks, Boulay Bay, Jersey, Mr. Larbalestier (May 1872). Spores “019 mm. long, "0085 mm. broad. No spermogonia detected. Medulla 1—. Gelatina hymenia I—. This very interesting lichen was sent to me for determination by Mr. Larbalestier immediately on its discovery in 1872; and I at once saw that it was a new and distinct species, and named it L. subdiluta, from the general external resemblance of the apothecia to those of L. diluta (Pers.). This name I communicated to Mr. Larbalestier (in litt.) in 1872 ; and he subsequently deposited specimens with that name in Herb. Brit. Mus. and in several private herbaria. Тһе lichen was subsequently shown by Mr. Lar- balestier to Dr. Nylander, who, as Mr. Larbalestier informed me (in litt.), thought it “а seaside anomalous state of L. prosechoides, Nyl., and not worth distinguishing," which of course it is not. Sincethen Dr. Nylander has changed his opinion, and now describes it as a “ species notabilis" in the * Flora,’ 1875, р. 442, under the name of Lecanora jejuna, Nyl. My name L. subdiluta, Leight., claims priority. LECIDEA ADVENULA, Leight. "Thallus obsolete; apothecia nigro-fuscous, minute, round, plane or subconvex, with a thin margin soon obliterated, brownish within; para- physes indistinct, conglutinate, apices large, nigro-fuscous ; hypothecium thickish, nigro-fuscous ; spores 8, nigro-fuscous, oblong, 1-septate. Parasitic on Pertusaria sulphurea (К. yellow, С orange-red). On rocks, Llanbedrog, near Pwllheli, North Wales, 1874. Very rare. A very singular lichen, not scattered uniformly over the matrix thallus, but occurring only on blackened patches formed by some conglomerate spherical brown alga at intervals over the thallus. Gelatina hymenia I—. Spores "024 mm. long, 018 mm. broad. MELASPILEA VERMIFERA, Leight. Thallus obsolete; apothecia black, very minute, irre- gularly angulari-oblong, imbedded in the cortical layer, when dry plane and sur- rounded by a minute upraised jagged margin of the cortical layer, when wet plano- convex, the margin obliterated; hymenium pale; paraphyses very delicate and slender, apices pale; asci lineari-obovate ; spores innumerable, colourless, arranged spirally in the ascus, cylindrico-fusiform, pointed and arcuate at each end, vermiform: 1-septate. Parasitic on the thallus and apothecia of Variolaria discoidea, Т. & B. (Pertusaria globulifera, Turn.). On oaks, Trefriw, North Wales, 1874. Rare. Gelatina hymenia I—. Spores "022 mm. long, "002 mm. broad. ARTHONIA PUNCTILLIFORMIS, Leight. Thallus a mere film; ardelle nigro-fuscous, exces sively minute, not larger than a pin’s point, irregularly roundish, convex, scattered, THE БЕУ. У. А. LEIGHTON ON NEW BRITISH LICHENS. 147 internally brown; asci pyriform ; paraphyses none; spores 8, pale fuscescent, ob- longo-clavate, 1-septate, large. On holly, Trefriw, North Wales, 1874, very rare. broad. Fig. 1. Spores “029 mm. long, "0135 mm. DESCRIPTION OF PLATE XXII. Verrucaria myriospora, nat. size. . Same lichen, enlarged. . Section of perithecium, Asci. . Spores, magnified 1200 times. . Spermogonia. . Spermatia, magnified 330 times. . Verrucaria succina, its apothecium and perithecium, magnified. . Apothecium as seen from above, showing papilla and epithecium. . Structure of perithecium. . Spore, magnified 1200 times. . Paraphyses. . Lecidea subdiluta, nat. size. . Apothecium, magnified to show its pecu- liar elevato-sessile character. Fig. 15. 16. 17. 18. 19. . Spores, magnified 1200 times. . Melaspilea vermifera: apothecia, mag- Section of apothecium. Spores, magnified 1200 times. Lecidea advenula, nat. size. Enlarged representation of the same. Section of apothecium. nified, as seen from above in dry state. . Apothecia wetted, magnified. . Paraphyses. . Ascus. . Spores, magnified 1200 times. . Arthonia punctilliformis, nat. size. . Ascus. . Spore, magnified 1200 times. Arihenia гигсааогуғы:з © [ 149 1 IX. On the Hygroscopic Mechanism by which certain Se $ are enabled to bury themselves in the Ground. By Francis DARWIN, M.B., F.L.S. (Рһлте XXIII.) п \ be | Read March 16, 1876. THE object of this paper is to describe the hygroscopic mechanism by which certain seeds are enabled to bury themselves in the ground. This phenomenon as it occurs in the Geraniaceze has already been described by Hanstein *, and again by G. Roux t in the case of Erodium; lastly, Asa Gray f briefly points out the power of burying them- selves possessed by the carpels of some of the Western American species of this genus. After describing the process, he adds, * It is the same with the grain and awn of Stipa.” I may mention that the latter note only appeared after I had completed my researches on this subject. The species оп which my observations have been made are Stipa pennata, Avena elatior, Heteropogon contortus (Bombay), Heteropogon (Andropogon) melanocarpus, (Florida), Androscepia arundinacea, Anthesteria ciliata $ (Khasia), Anemone montana (from Switzerland). 4 It seems to me an extremely remarkable fact that the curious power of burying them- selves should be exhibited by plants belonging to several distinct orders, and inhabiting various parts of the world; and that the mechanism should be essentially the same in all of them is even more remarkable. My observations have been chiefly confined to Stipa; a short description of the caryopsis of this plant will explain the general method of action common to all the other fruits observed by me. Fig. 1 represents the seed || of Stipa pennata. It terminates inferiorly in a sharp strong oblique point (р) €, armed with a dense plume of barb-like hairs; v, Ё. fis the strong woody awn, of which the lower vertical part (v) is strongly twisted on its own axis, the appearance of the strands of a rope being given by the ribs r and ғ, fig. 9 (and seen in section in figs. 11 & 12), which run spirally up the awn. Above, the awn terminates in a long untwisted portion (f), bearing a series of beautiful hairs, giving its well-known feathery aspect: when the seed is dry this portion extends nearly horizontally out wards. The twisted and vertical part (v) is separated from the feathery part (f) by а curious double bend; the two angles thus formed I call the lower and the upper knee (4, and k»). * Sitzungsberichte d. Niederrhein. Gesell. Bonn, 1868. See Sachs, * Botany, "Eng. Trans. p. 841, for abstract. + Annales de la Soc. Bot. de Lyon, 1873. t Silliman's Journal, Feb. 1876, p. 158. : $ Dr. Hooker most kindly supplied me with specimens of the four last-mentioned species. | | The use of this expression in place of the strict botanical term шау perhaps be allowed me, on account of its Breater convenience. т : *| Said to be eapable of injuring the intestines of animals which feed on the grass (Hooker's Trans. of Decaisne and Le Maout, p. 892). 159 DR. FRANCIS DARWIN ON THE MECHANISM BY WHICH The hygroscopic action of the awn may be shown by holding the seed so that the twisted axis is vertical; on wetting the awn, the feathery portion will revolve, describing a circle in a horizontal plane; the vertical part untwisting, and its strand-like ribs becoming more and more oblique ($. e. making a smaller angle with the axis of the awn), until at ‘last, when the awn is thoroughly wet, it shows no signs of torsion, and the two ribs run parallel to its axis down two of its opposite surfaces. Аз the awn untwists, the two knees, А, and k, are obliterated (k, disappearing first); by this means the feather is brought into line with the twisting part, and the awn forms a long straight pliable rod, pointing vertically upwards. It follows from this, that though the end of the feather begins (at the commencement of the wetting) by describing a circle, it ultimately describes a helix. As the awn dries again, the action is exactly reversed : the axis becomes twisted; the two knees reappear ; and the feather becomes again subhorizontal. It was this remarkable property exhibited by Stipa of untwisting and twisting in response to alternations of moisture and drought, which convinced my father and myself that the whole structure must be adapted to some very definite end. We concluded that this object must be that of forcing the seed into the ground (in the manner to be imme- diately described). The sequel gives the result of my investigation of this point. If in the above-described experiment the feather is prevented from revolving while the seed is left free, the untwisting of the vertical part of the awn will make the seed rotate on its axis. Іп consequence of the untwisting, the vertical part of the awn from the knee (2,) to the tip of the seed becomes longer by about 7 per cent. Therefore, if we fix a dry Stipa-awn іп a vertical position, the seed resting freely on the ground, and the feather being prevented from revolving, on wetting the awn the seed will rotate and will be pressed against the ground. But this requires that the expansion of the awn shall not be permitted to take place in an upward direction; i.e. the upper extremity must be fixed to givea point d'appui, from which the expansion may take effect in pressing the lower extremity against the soil. In describing the untwisting of the axis it was stated that the angles k, and 4; increase as the амп absorbs water, until the whole awn is converted into а straight rod. Now in the same way that, by interfering with its movements, the rotation of the feather is transferred to the seed, so by preventing the feather from rising to its full extent the tendency of theawn to become straight results in increased pressure of the point against the ground. Тһе seed now resembles a brad-awl in exhibiting vertical pressure оп 4 revolving point. In order that this action may take place in its entirety, three poinls d'appui are theoretically needed, although one is practically capable of performing the duties of all. The first is required, to transfer the rotation from the feather to the seed ; the second, to force the expansion resulting from the untwisting of the awn to take effect downwards; the third, to convert the tendency of the awn to straighten itself into pressure of the seed against the ground. It is evident that if an awn be placed vertically, the seed resting on the ground, so that the feather can neither rotate nor describe any great angle in a vertical plane, all the required conditions will be fulfilled. How can this be effected in a state of nature? Prof. Hildebrand has pointed CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 151 out * that the long feathery awn of Stipa serves as a “ Verbreitungsmittel," or means of distribution. It is found that seeds dropped from a height of a few feet usually preserve a nearly vertical position, and strike the ground with their point. Moreover, if they are allowed to fall among low vegetation they are caught by the knees (2, 4) or by the feather, and are fixed in a more or less oblique position, the seed resting on the ground. Many seeds would no doubt escape being entangled; and many others would be caught in unsuitable positions—suspended in the air, for instance, or making too acute an angle with the ground, or resting on some hard impenetrable body. Neverthelsss many seeds are undoubtedly caught and held fast in suitable positions. The great length of the feather (nearly 30 cm.) would render entanglement in low vegetation an easy matter; and when a seed is once entangled the hairs serve to hold it fast and prevent the wind blowing it away. In order to ascertain what would occur under these circumstances, I made experiments of the following kind :—A vessel being filled with dry sand or light soil, two upright sticks were fixed at about a centimetre apart; and these were connected at the height of 8 or 9 сш. from the surface of the soil by a cross pin, a second one being fixed 2 ог 3 cm. higher up. The Stipa-seed is then placed with its point resting on the soil, and its feathery awn passed through the loophole between the two supports and the two cross pins. The whole is now placed under a bell glass lined with wet blotting-paper. This is better than wetting the awn directly; for in touching it with a wet brush it is possible to push it into the soil without being aware of it, and in syringing it the soil may become too much wetted. In order to dry the awn it is merely necessary to remove the glass; thus by alternations of dryness and moisture the awn is made to twist and untwist, just as it would under the changing hygrometric conditions of dew and sunshine in a state of nature. The amount buried may then be measured with a pair of compasses. It is very important to estimate the amounts buried, by measuring from the surface of the soil to the top of the seed, and not to any marked point in the амп; for in the latter case the measurements will be obviously inaccurate, owing to the lengthening and shortening of the awn. All the movements of the awn are far more conveniently studied when the seed is fixed and the feather free to rotate; and in taking the times of rotations &е. I have found it best always to employ immersion in water in preference to exposing the awn to a damp atmosphere; for by the former means a constant amount of moisture is secured, and in nature the awn must be frequently exposed to an amount of moisture practically equivalent to immersion. A dry Stipa-awn plunged into water at the temperature of the room began to rotate within a second of its immersion, and untwisted at the rate given in the fol- lowing table :— * Verbreitungsmittel der Gramineen-Früchte, Bot. Zeit. 1872, No. 49. SECOND SERIES.—BOTANY, VOL. І. 152 DR. FRANCIS DARWIN ON THE MECHANISM BY WHICH Turn. | Completed in | Turn. Completed in | M. 8. | м. 8. | No. 1 Sw Ио 6 130 -—| 2 9:20 | 7 1 45 3 1 45 | 8 2 10 4 id 9 3 90 5 19571 10 | - Thus the rate increases up to the 5th revolution, and then diminishes quickly. The lower bend or knee is obliterated in from 20’ to 30’; the awn becomes com- pletely untwisted in from 3 to 1 hour; in doing so it increases in length by about 7 per cent.; a well-grown awn measures from the point of the seed to this lower bend 70 mm. dry, 75 wet. The upper knee disappears in periods of time which are rather variable, and lie between two and three hours. The movement which takes place as the awn untwists or twists is a combination of two rotations: the first is that due to the torsion of the vertical portion of the awn (50-55 mm.) from the lower knee down to the seed; the other rotation is due to the torsion of the short portion (15-18 mm.) between the knees. This latter axis is carried round quite independently of its own torsion, by the torsion of the long vertical axis; thus the portion Ё, k, (fig. 1) describes a cone about an axis of rotation which is continuous with v; and, again, owing to the torsion of the part kı b, the feather is made to rotate, describing a figure of rotation having a line con- tinuous with А, 0; for its axis. We have seen that the feather is usually nearly hori- zontal when dry, and vertical when wet (the seed being held vertically in both cases); it follows that in passing from the wet to the dry state the tip of the feather does not describe а regular helix, as would be the case if there were only one bend, but a complex figure varying with the varying relations between the rotations due to the two axes 2 and А, Ё. The rotations due to tlie untwisting of Ё, Ё, are slow (7-18 minutes) ; and since but two revolutions are possible, owing to the shortness of this part, their effect is not always great on the revolutions due to the untwisting of the vertical axis, which are from 13 to 16 in number. Тһе rotation of the алуп is best studied by fixing the seed while the upper end is allowed to revolve; and by the experience gained in this way it is possible to understand what occurs in the experiments on artificial burying where the awn is fixed while the seed is free to rotate. When an awn has been placed for seyeral hours in the damp chamber, fixed in the kind of support described, it will not always be found to have buried itself. I have often found that the point of the seed has been dragged across the soil into which it should have thrust itself; and in these cases it has . happened that the seed has been pushed down over the edge of the vessel, so that it would have tended to bury itself if the soil had been of a larger area. It is this dragging motion which I believe to be produced by the torsion of the short axis between kı and №, (fig. 1); for just as, when the awn is free, irregular movements are impressed on е rotation, so ће tendency to irregular rotation is transferred to the seed when ће awn is fixed. In other cases the experiment is successful, and the seed is found buried to а certain depth; the sand is often found disturbed close around the point, owing, DO | CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 153 doubt, to the irregularity of the rotatory motion of the awn. Тһе following measure- ments give the amounts buried in three cases during a single wetting :— $: З Р.м.—А seed of 16 mm. length, resting with its point on damp sand. 9 P.M.—14 mm. of seed projecting above soil. 9 л.м. (still wet).—10 mm. ,, J Amount buried 6 mm. (2 of them in 2 hours). 11. 5 Р.М,--12 mm. of seed projecting. 9 A.M.— 7 mm. b á Amount buried 5 mm. AH. 11.0 А.м.—6 mm. of seed projecting. 3.15 4M.—2 mm. ,, » Amount buried 4 тит. On removing the bell glass—that is, on substituting a comparatively dry for a damp atmosphere—the awn begins at once, 4. е. in less than one second, to twist in the oppo- site direction to the hands of a watch; and as the drying proceeds all the above-men- tioned movements are reversed. Тһе following are the times in which the first six revolutions were made in once instance :— | un Completed i in | 'Turn. Completed i in | м. 8 | Ж Е | Мела 1902204 No. 4 280 | | 2 9:40 | 5 6. D | | 9 g U | 6 7 30 | | т з, the movement being slow at first, then becoming quicker, and then slow again. The lower knee (/,) becomes perceptible in from 5 to 10 minutes, the e ior knee (ka) begins to be perceptible in from 10 to 15 minutes. Тһе whole рсете of drying — be said to be thoroughly complete under two hours (in ап ordinary e TAs but the movement during the second hour is extremely slow, and practically almost unimportant. : с When the drying-process is finished an examination of the seed will show that опе of three things has occurred :— : (i.) The upward traction which results from the return of the awn to its former wid condition frequently pulls the seed out of the soil, especially if the sand used in the experiment is dry. 5 (ii) The seed may, however, be able to withstand the upward traction. Qiii.) It may actually be found to be buried deeper than it was at the commencement _ Of the drying-process. 4 Y 154 DR. FRANCIS DARWIN ОМ THE MECHANISM ВУ WHICH The following measurements give the amount buried in three instances during a single drying :— L 6.45 P.M.—9 mm. of seed projecting above surface. 8.15 А.м.—7 mm. 5 Amount buried 2 mm. ә 15% 11.30 А.М.- 192 mm. projecting above surface. 1. 2 р.м.—10 mm. о b Amount buried 24 mm. TIE 9 A.M.—10 mm. projecting above surface. 11 А.М.-- 6 mm. 33 РЕД Amount buried 4 man. In investigating the burial of the seed caused by drying, it must first be remarked that the plume of hairs with which the point is armed (see figs. 1 & 7) offers but little resist- ance to the entrance of the seed, but tends to prevent its withdrawal. Тһе bare sharp point ( p, figs. 1 & 7) is about 3 mm. in length; the hairs which form the plume are graduated, the smallest (1 mm.) being the nearest to the point, so that they tend to preserve the smallest amounts buried during the wetting process. Moreover the length of the whole plume, measured from the extreme tip of the seed, is only about 4 mm. ; and as we have seen that as much as 5 mm. may be buried in one wetting, it is quite con- ceivable that the seed can resist the uprooting tendency of the drying process. То understand how it can actually be buried deeper, we must examine more closely the act of rotation during drying. А vertical rod, which is fixed at the upper extremity, will as it twists on its axis, cause a simple rotation of its lower free extremity. But if the upper or fixed end is not vertically above the lower or revolving end, 4. e. if the rod is curved, there will, 1 believe, be a tendency for the rotation to take place about the line joining the upper and F lower ends of the rod. The movement imparted to the point will then resemble the circular rocking motion which is employed when а stick is being thrust into the ground. Given pressure from above, a rotating vertical rod will make a cylindrical hole of continually increasing depth; but if this pressure is removed, it will merely 7S continue to rotate in the hole already made. But in the case of à g n: 4 Stipa-awn (a curved rod twisting on its axis), even when the pres- P sure from above is removed the rocking* tends to continue the bury- ing process. Тһе point of the seed may be compared to a barbed point, like that in the woodeut. The"point P is supposed to be thrust into the ground, gg’ giving the level of the soil. * Whatever may be the explanation of this movement, it certainly occurs. CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 155 The free end Е is now supposed to rock backwards and forwards, describing ares alter- nately in directions $ & i 7, in the plane of the paper and about the point P. When Е moves in direction ¿ the barbs ddd will offer little resistance, whereas the barbs cee will in fact act like fulera, and will enable P to be forced deeper into the ground. On the movement being reversed, this action will be reversed also, 444 giving the fulera, and P being again buried deeper. Although the tendency of а Stipa-awn, on drying, is to move (not backwards and forwards in one plane, but) so as to describe a cone, yet I believe the above-described process is essentially what occurs in Stipa; for the seed is revolving in soil of densities varying at different points, so that it moves not regularly, but by jerks; moreover the plume is rather more developed on two of the opposite surfaces of the seed than on the intermediate regions, and the point is obliquely set оп. In observing the process of drying, I have often seen a distinct rocking move- ment as the awn attempted to rotate and was continually forced back again. All their aberrations from a structure and from a rotation of mathematical regularity must favour the levering or wriggling movement of which the diagram in the woodcut gives the essential action. In describing the manner in which the seed of Hrodiwm is buried during drying, Han- stein does not enter into any of the above considerations as to a rocking movement &c. ; he merely says that * the lower part of the awn begins to contract into narrow spiral coils, causing the cone (i. e. seed) to turn on its axis and penetrate the ground; and the егесі hairs on it, which point upwards, retain it there like grappling-hooks ”%, It has now been shown by what means the seed of Stipa is buried, both as it untwists and also as it returns to a state of torsion. Ву a combination of these two processes, the awn is thrust into the soil to such a depth as to cover up the seed completely. Thus, in an experiment, in three wettings and three dryings 28 mm. was buried in dry sand: the ordinary length of a seed is 17 mm.; so that this is amply sufficient. In another experi- ment a seed of 16 mm. length was completely buried in three wettings and three dryings. Another seed, which I entangled in the branches of a low bush, and left out of doors for eight days, had buried itself to a depth of 31 mm., impaling a piece of rotten leaf on its way. Mr. Farrer informs me that the Stipa-seeds which blow away from the parent plants succeed in burying themselves in his garden. The question of what advantage it is to the plant to bury its seeds will be discussed in the sequel. Before passing on to describe the arrangement of structure of which the hygroscopic mechanism consists, I shall give a brief account of the phenomena exhibited by the Stipa-awn, considered merely as mechanical actions, and with no reference to the biolo- gical conditions in which they occur in Nature. A simple instrument was made for me by my brother Horace: a piece of Stipa-awn, about 5 em. in length, was employed, taking care to avoid either of the knees (k, ог kə). The awn is fixed at one end, and at the other bears a light index, fitted to travel round a clock-face as the awn twists or untwists; the bearing of the awn is so managed that the whole of it can be immersed in different fluids. The S¢ipa-awn is hygroscopically ex- * Sachs, p. 841, Eng. Transl. (a perfectly correct abstract of the original). 156 DR. FRANCIS DARWIN ON THE MECHANISM BY WHICH tremely sensitive *; but as we shall see that it is also very sensitive to changes of tempe- rature, a purely hygroscopic result is not so strikingly demonstrable as one due to a change in temperature and in moisture combined. Thus, breathing on the Stipa-hygroscope, held even at arm's length, or bringing the warm and moist hand near the awn, causes distinct movements of the index. De Lue showed t that water has no special virtue in causing the expansion of hygro- scopie bodies, and that other fluids can perform the office of filling out the molecular interstices. If the Stipa-hygroscope is allowed to dry in the air of an ordinary room, and is then plunged into absolute alcohol, the index shows, by moving in the “ dry” direction, that water is being removed from the awn; after a time the alcohol itself is absorbed, and the movement of the index is reversed. On removing the instrument, the index moves in the “wet” direction, in consequence, І suppose, of the absorption of water by the alcohol; ultimately, however, the alcohol evaporates, and the movement of the index is reversed. The effects of temperature on the awn are extremely curious, and agree with those observed by De Luc 1: he found that a rise in temperature affects hygroscopic bodies in the same way as increased moisture, whereas a fall in temperature acts like dryness. With a Stipa-hygroscope the experiment is a pretty one; it is allowed to remain in warm water until the index comes to rest; if it is then removed, and plunged into cold water, the index gives a quick start through 90° or more in the direction of drying (2.6. against the hands of a watch) On replacing it in the warm water a rapid movement in the opposite, or “ wet," direction takes place. Тһе experiment may be performed in another way. Ifthe hygroscope is placed in the current of hot air from a lamp, the rise in tem- perature tends to make the index move in the wet direction; but it also dries it, and therefore tends to make it rotate in the opposite direction: the struggle is made obvious by the fluttering motion of the index. Тһе ultimate result is the victory of the * drying" rotation. The coincidence in the effects of heat and moisture accounts for the extreme sensitiveness of the S¢ipa-hygroscope to being breathed on. The first theory which suggests itself to account for the effects of temperature is, that as an increase in the quantity of water in the molecular interstices makes the awn untwist, an increase in bulk of the water already permeating the tissues acts in the same manner. In support of this view my brother Horace showed that it cannot be due to air contained in the tissues; for the index does not move when the receiver of an air- pump under which it is placed is exhausted, or when the air is readmitted. In per forming this experiment we immersed the hygroscope in oil or in mercury, to obviate the effects of the change of temperature resulting from the change of air pressure; before this precaution was taken the index moved distinetly as the air was removed or let in, although the thermometer only registered a change of 2° F.$ * It has been employed as a hygrometer (Watts, Dict. of Chemistry, iii. p. 233). + Phil. Trans. 1791, p. 11. t Loc. cit. p. 16. $ The sensitiveness to changes in temperature was well exhibited by my brother, who observed that on lifting the Stipa-hygroscope out of water it always moved very slightly in the cold direction, but that this movement disappeared at the temperature of the dew-point, showing that it was due to evaporation. CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 157 My brother then suggested that if it were due to the expansion or contraction of the water contained in the tissues, the results ought to be exactly reversed when water at 4° C. was used as the higher temperature, 0° being the lower one, because water expands instead of contracting in passing from 4° to freezing-point, and therefore the Stipa-hygro- : scope ought to behave as if 0° was a higher temperature than 4°. Preliminary experi- ments showed that a difference of 4^ is clearly indicated by the hygroscope; but, to our surprise, we found that the hygroscope behaves exactly as if water expanded in passing from 0 to 4°. We therefore concluded, like De Luc, that the effects of temperature have nothing to do with the expansion of water. He placed his hygrometer in quicklime, by which means he obtains what he calls * absolute dryness," and found that changes of temperature affected his instrument in ** nearly the ваше” manner as when in water. The only explanation left appears to be that the woody tissue itself expands with heat. But, from the experiments of Villari *, it appears that the expansion of dry wood with heat is extremely small compared with the expansion due to imbibition. I find it impossible to believe that the rapid rotation through a considerable angle is due to this cause. Тһе following eurious phenomenon negatives such a view, but must also remain quite unex- plained. It will be made clear by detailing an experiment. Two vessels of water were employed, whose temperatures differed by about 68° C. In the cold water the index of the hygroscope stood at 130; on putting it into the hot water the index moved quickly to about 240 (not exactly noted), and returned slowly to 115. ‘That is, it first makes а rapid wntwisting, or wet movement, and then returns to a point on the dry side of its original position—that is, a point representing increased torsion of the awn; so that on moving into hot water it untwists rapidly, and then twists slowly up beyond its original degree of torsion. Exactly similar but reverse results ensue on removing the awn from hot to cold water. These experiments have been frequently repeated by my brother and by myself. The Mechanism of Torsion.—In Sachs's * Handbook’ t the twisted growth of trees is explained by the internal parts not growing as fast as the external tissues, the unequal longitudinal tensions satisfying themselves by producing torsion. To apply this to the tor- sion of the SZipa-awn, we must suppose that, on drying, the internal parts contract more strongly than the external. "This action may be imitated, as Sachs remarks, by slipping an elastic tube over a second of smaller calibre; the internal one is then stretched, ihe external one being left as it is, and the two are bound together in several places. On the release of the internal tube, the external one will not permit any shortening of the internal tube except by the whole system turning into an irregular helix. Hil- debrand { offers a similar explanation for the torsion of the awn of the * Springhafer ” (Avena sterilis). He says that one surface of the awn contracts on drying more strongly than the opposite surface, and the awn twists on itself to satisfy the unequal tension thus produced. Hanstein $ explains the torsion of the awns of the Geraniacez ina Similar way. For a long time I concluded that some explanation of a kindred nature * Quoted in Sachs: English танба, р. 649. + English Translation, 1874, р. 770. t Pringsheim's Jahrb. für w. Bot. 1873, Bd. ix. “ Sehleuderfrüchte." $ Loc. cit. 158 DR. FRANCIS DARWIN ОХ THE MECHANISM BY WHICH must hold good for Stipa. I assumed that if there existed unequal longitudinal tensions along opposite surfaces of the awn, there would be inequality of contraction or expansion in a radial direction, and that the sectional outline would alter in shape on drying. I could observe no such alteration, and found, in fact, that on swelling in water the awn increased in diameter very nearly equally in all directions. І con- cluded, therefore, that the inequality of expansion must be between the internal and external tissues, and not between two longitudinal halves. I saw no way of con- firming or destroying this hypothesis; but I always felt a difficulty in the fact that the Stipa-awn is always twisted іп one direction, whereas the above-mentioned unequal distribution of tensions would give a tendency to twist in either direction. Ulti- mately all theories of unequal contraction of the awn as a whole were overturned by observing that longitudinal sections and mere strips torn from the awn were capable of twisting up into precisely the same right-handed screw as the whole awn. ‘This proved that torsion-power may reside in а combination of a few cells. The fact that a transverse section of the twisted awn of Anemone montana shows only broken- down tissue in the centre, convinced me that torsion is possible where the twisting- organ has no strongly contracting tissue in the centre. These two observations suggested that the power of torsion must reside in the individual cells composing the awn. I therefore boiled a piece of Stipa-awn in dilute nitric acid and chlorate of potassium; by subsequent teasing in a drop of water, the cells were isolated with great ease. The slip of glass was then held over a spirit-lamp, and a light teasing action continued, to prevent the cells adhering to the glass on drying. On examining the object under the microscope the torsion of the individual cells was beautifully seen; besides individual cells, numerous little ropes of two or three cells are seen, all twisted in the direction of torsion of the awn, i.e. that of a right-handed screw (see fig. 14). To this observation it may be objected that the treatment with hot acid may have conferred the power of torsion on the cells. То this it may be answered, (1) that small portions, separated by mechanical means, exhibit torsion when dried; (2) that pieces of awn treated with boiling nitric acid and chlorate of potassium (if they are not teased into their constituent cells) undergo torsion on drying, proving that the cause of the torsion, whatever it be, is not affected by the treatment with acid. I believe that both the internal and the external cells of the awn are capable of inde- pendent torsion; but whether this be so or not, it is certain that the small external cells exhibit the power in a far higher degree. "This cleared up what had always seemed à great difficulty, viz. the instantaneous movements of the Stipa-hygroscope ; for it removes the seat of hygroscopic action from the internal tissues to a more accessible part near the surface. In accordance with the present theory, it is found that the individual cells possess the same delicacy of action as the whole awn. I have actually seen a single cell under the microscope untwisting and twisting up again as my hand approached and was withdrawn from its neighbourhood. Supposmg ы then to be granted that the cells composing the awn are capable of inde- pendent torsion, we have yet to show that the twisting of the awn as a whole will be - CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 159 the result. The cells are of an elongated form, and are of course closely attached to one another, so that unless they are isolated they cannot twist on their own axis, since this would require a sliding movement between each cell and its neighbours. "The following experiment demonstrates the way in which each cell satisfies its tendency to twist on its own axis. А number of S/ipa-awns (exclusive of feather and knees) are soaked in water, and when thoroughly untwisted are firmly tied into a cylindrical bundle or faggot. Here we have represented the state of things in an awn: each constituent awn in the bundle represents one of the cells of an awn, and, like them, tends to twist on its own axis on drying. It is found that, on drying, the bundle of awns is converted into a rope, its constituents passing helically round like the strands. And just in the same way the cells which make up ап awn (of which the bundle of awns is a schema) satisfy their tendeney to twist by forming a rope of cells twisted in the same direction as themselves. If, then, we have a number of cells so connected together that they are incapable of independent torsion (and this is the case in the awn of Stipa), and if, further, each cell has a tendency to twist on its own axis (as we know to be the case with Stipa), then it cannot be doubted that the torsion of the mass of cells as a whole will be the result. That this is actually the mechanism by which the awn twists is conclusively shown by the fact that the constituent cells of the feathery non-twisting portion have ло such power of independent torsion when isolated and dried in the manner already described. The Bending of the Awn at the Knees (k, & k.).—1f we compare a section of the feather- bearing portion of the awn with a section of the twisting part, we find that the difference between them lies in this: in the twisting part the cells are all thick-walled (excepting the central ones, с, and two small masses, m & m, fig. 11); but in the feather all the cells are hollow, and it looks as if the central mass of thin-walled cells had increased so as to fill up the whole interior of the section. And since this thin-walled tissue is not hygroscopic in the feathery portion, we may assume that it is not so in the twisting part. Comparing a section taken low down in the awn (fig. 11) with one taken at the upper knee (fig. 13)*, we find a marked difference in the distribution of the non-hygroscopic tissue, At the upper knee the cells are passing into the hollow condition found in the feather; but this change does not attack all parts simultaneously: the small masses of non-hygroscopic tissue, m & т, have enlarged and coalesced with the central mass (с); and it is the cells surrounding the coalesced masses which exhibit the commencement of the loss of thickening in their walls. We may divide the section by a line гу, on one side of which there are many hollow cells, on the other all thick-walled cells. The twisting-power, which is already weak in the portion of the awn between the two knees, must be disappearing at k», since in the feather it is quite gone. When drying com- mences, the mass of thick-walled cells contract longitudinally ; and the mass of hollow cells not being able to contract to the same extent, the awn will bend with the non- * In the drawings of the sections, figs. 11, 12, & 13, the stratification of the cell-walls is not represented; nor are the pit-channels given, which are especially numerous in the large internal cells, and which seem to communicato With the vascular bundle in the centre. SECOND SERIES.—BOTANY, VOL. I. Z 160 DR. FRANCIS DARWIN ON THE MECHANISM BY WHICH hygroscopic half on the convex surface, and the strongly contracting cells on the concave aspect. То prove that the internal cells do contract longitudinally, it is only necessary to split an awn longitudinally ; on drying, the split-off pieces bend with the internal surface concave. There can be no inherent tendency in the separate cells at 0; to bend in the direction assumed by the awn; for a small portion severed by a longitudinal section (such as is shown by the upper dotted line, fig. 13) bends with the internal surface con- cave, which is in the opposite direction to that in which the whole awn bends. The bending at the lower knee is more difficult to explain. The hollowing-out of the cells near m & m has begun even at this part of the awn: two hollow cells are shown in fig. 12; and in other sections I have found more of them occurring. Іп this section, again, we might draw a line which would divide the awn into a more and a less contractile half. In fact we have at the lower knee a bending-mechanism similar to that which exists at the upper knee, but with an important difference: at k, the torsion-power of the awn is exhausted, whereas at the lower knee this is not the case; and since the unequal longitudinal tension just described can satisfy itself by assisting the torsion, it does not seem evident how the bending of the awn is to be effected. To explain this a hitherto neglected point in the structure must be noticed. If we compare once more a section of the awn close to the seed with one taken at either of the two knees, we find that in the lower section (fig. 11) the ribs ^^, are separated from one another by approxi- mately equal portions of the circumference; but at the lower knee (fig. 12) the are on the side 2% m has diminished, while the other has increased. In a wet or untwisted awn, the fact that the ribs approach one another as they pass from below upwards, gives the appearance of a very slight and elongated left-handed screw. And I believe that the ap- proach of the ribs to one another is directly connected with the unequal arrangement of the more and less hygroscopic tissue at the knees. At the lower knee there is, moreover, a single sudden and short turn of a left-handed screw. This I call the ** reverse twist," since it is in the opposite direction to that in which the awn twists in drying. The reverse twist is perfectly distinct from the ordinary hygroscopic torsion of the awn ; for it is not obliterated by wetting, and is the result of the growth of the tissue into that shape. It is like a single turn of a fluted spiral column made of india-rubber, which could obviously be twisted in either direction independently of its shape. Now since the awn twists in the opposite direction to the reverse twist, the latter will be obliterated аз torsion proceeds ; and this is, in fact, the case: the awn is twisted both above and below the knee; but at the actual bending-place there is no torsion. When the reverse twist is obliterated, the unequal longitudinal tensions at the point б; are employed in producing the bend, instead of satisfying themselves in assisting the torsion. ‘To summarize this imperfeet explanation, the torsion-power satisfies itself in obliterating the reverse twist; and then the unequal tension, being brought into a longitudinal direction, satisfies itself by producing the bend. I suspect it was the reverse twist which deceived Max Wichura*, and led him to de- scribe the awn of Stipa as twisted to the right below the knee, and to the left above it. г * “On the winding of Leaves,” Taylors ‘Scientific Memoirs; May and Aug. 1853, p. 280; the original ™ * Flora,' 1852, Jan. and Feb. CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 161 I have diseussed the mechanism of the bend at some length, because of its extremo importance in the burying-process; for if ће awn did not terminate above in a non- twisting horizontal portion, the rotation would not be transmitted to the point of the seed, neither could there be any increase of vertical pressure resulting from the straight- ening of the awn. Torsion.—An explanation will now be attempted of the power of torsion which the cells possess. It is a general property of hygroscopie tissue that the cells composing it are thick- walled. I have already pointed out the difference in this respect which exists between the hygroscopic and the non-hygroscopic tissue of the Stipa-awn. Hildebrand (* Schleuder- früchte ”) mentions the thick-walled cells composing the awn of Avena. We should expect therefore that the power of torsion would depend on the manner in which the cell-walls are thickened. Before discussing my view, it will be well, first of all, to exclude any cause resembling the unequal internal and external tensions which are supposed to account for the torsion of trees; for this only shifts the difficulty one step further off, and does not account for the constancy of direction of torsion. We may now examine the structure of the cells which make up the hygroscopic tissue. А body which swells on imbibing water will expand equally in all directions, if the molecular interstices in which the water lies are symmetrically arranged in all directions. Anda cell will have no tendency to twist if its cell-wall expands equally in all directions. Therefore, if we are to account for the torsion of a cell by the expansion or contraction of its cell-wall, we must examine the molecular structure of the wall with special reference to the distribution of the capacity for absorbing water. Тһе well-known researches of Nageli* are directed to this very point. He there shows that the cell-wall is composed of parallel lamell:e of alternate degrees of density and refractive indices. The first series of lamellz are seen in transverse sections of elongated woody cells, as concentrie shells, alternately light and dark, and fitting inside each other. This appearance is well known as stratification or ** Schichtung.” Тһе other systems are essentially of the same nature, but are not so well marked. They give rise to the appearances known as “ Streifung " or striation ; these are series of parallel lines, alternately light and dark, traversing the surface of the cell, and are in reality the edges of parallel lamelle of alternate densities. There are usually two systems of parallel lamelle ; and they may be inclined to the axis of the cell at almost any angle. Very frequently the two systems wind spirally round the axis in opposite directions. Now according to Hofmeister +, when the tissue of the cell-wall expands during imbibition, it is chiefly due to the swelling of the less-dense striæ; and we have seen that these stris are spirally arranged ; therefore we are led to expect that the imbibition of water will result in some kind of spiral tension: and spiral tension will result in torsion—just as when a string is fastened to one end of a rod, and is coiled spirally round it, and the free end is pulled, the rod will tend to rotate on its axis. And since there are two systems of spiral striation, the tension due to one system must be stronger than that t Lehre v. d. Pflanzenzelle, (1867) p. 197. * Münchener Sitzungsb. 1864, May & July. : 272 162 DR. FRANCIS DARWIN ОМ THE MECHANISM BY WHICH due to the other, if any tendency of the cell to twist in one direction more than in the other is to result. Nägeli has shown * that imbibition does lead to the torsion of a cell on its axis when one system of striation is more pronounced than the opposite. Fig. 10 is copied from his paper already quoted t: а shows a cotton-wool fibre in extremely dilute sulphuric acid; 0 is the same fibre in somewhat more concentrated acid. The first treatment merely brings out the stri] more clearly; the stronger acid causes the fibre to swell and to twist on its axis. Тһе importance of these figures is this: in а 16 may be seen that the direction of the striation changes in the lower half of the fibre; and in b it is seen that the direction of the torsion changes precisely where the direction of the striation does; not only do cotton-wool fibres twist on the intense imbibition caused by acid, but, as I find, on the intense abstraction of water caused by strong drying. From these facts we must conclude that the striation of the cells is the cause of the torsion of fibres of cotton wool. The elongated cells of the 5Нра-алуп present a close analogy with cotton-wool fibres; not only do they twist on drying, but also, as a result of the great swelling caused by macera- tion in strong Schultz’s solution, they undergo torsion in the opposite direction. From all these considerations, I cannot resist the conclusion that the torsion of the cells in the Stipa-awn is a direct consequence of unequal contraction of the cell-wall due to the striation of the membrane. I have found the investigation of the molecular structure of the cells in the Stipa-awn too difficult a task to be included in the pre- sent research; I believe, however, I may say that the cells are obliquely and spirally striated, and that one system is more strongly developed than the opposite. Both МасеН and Hofmeister give explanations of the ultimate mechanism of the tor- sion of cells; but in my present need of clear anatomical details it would be useless to apply their explanations to Stipa-cells. I now pass on to describe the remaining seeds or fruits observed by me. AVENA ELATIOR. Professor Hildebrand has described the hygroscopic awn of Avena sterilis, both as to the mechanism of the twist and the adaptation of the hygroscopic torsion as a means of distribution f. Fig. 4 shows the awn of Avena elatior (the empty glumes having been removed). It will be seen that, supposing the fruit to be held vertically, we have, just as in Stipa, а vertical twisted part of the awn, and а more or less horizontal part which is not twisted. In Avena sterilis, according to Hildebrand$, the spikelet contains two fertile flowers, and there- fore two bent awns. Не describes how, in drying, the pressure of the rotating part of the awn against the ground causes the fruit to be projected into the air (“ein Stück . fortgeschleudert wird ")|. * Münchener Sitzungsb. 1864, July, p. 124. T Tab. i. figs. 8 & 9. + Pringsheim's Jahrb. loc. cit, ; and * Verbreitungsmittel der Gramineen-Früchte," Bot. Zeitung, no. 49, 1872. $ Schleuderfrüchte, Separatabd. p. 14. | This curious property has gained for it in Germany the popular name of “ Springhafer " (Hildebrand). I have met with a eurious mention of the wild oat in the * The Young Man's Companion, or Arithmetic made easy, '1721:— CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 163 He also describes a method of progression by turning over and over, and a kind of creeping movement caused by the extension and flexion of the awns. Prof. Hildebrand suggests that possibly the more important function of the twisted and bent awns of - Avena and other grasses may be similar to that of Erodium, i. e. to bury the seeds in the ground. From analogy with Stipa, and from other considerations, there can be little doubt that this surmise is correct. | The fruit of Avena elatior, 6 times magnified, is given in fig. 4. Тһе vertical part of the awn (attached to the outer palea or flowering glume) is about 5 mm. in length, and is strongly twisted into a right-handed screw of about 4 turns. It follows that, justas in Stipa, the horizontal portion (6 mm. long) revolves in the direction of the hands of a watch when the awn is wetted. It is more sensitive hygroscopically than Stipa, and untwists very quickly (one turn in from 15” to 30”) when placed in water, as compared with the Stipa-hygroscope, which makes one revolution in 2 or 3 minutes. It will be seen that the point is blunt, and that it is covered with a plume of hairs pointing backwards as in Stipa; but these hairs seem to me too weak to be of much service. The horizontal portion of the awn is not feathered, but armed (as is also the twisting part) with minute reflexed barbs. I have not succeeded in observing the process of burial in Avena; nor has Hildebrand seen it. But I found that the seeds of a patch of wild oats growing in a ploughed field had succeeded in burying themselves. I believe that the mechanism of burial is not quite the same in Avena as in Stipa; the seed is too heavy to be held vertically, and the awn does not seem fitted for so supporting it. More- over the specimens which I found buried were not in a situation where low herbage could have entangled their awns, but among bare lumps of clay. I presume that the point is pushed laterally against inequalities in the ground, other projections being made use of as points d'appui. In explaining the mechanism of torsion in Avena, I am compelled, unfortunately, to differ entirely from Prof. Hildebrand. His view has been already mentioned. Against it I bring forward the arguments :— (i.) That it does not account for the direction of torsion being constantly the same. (ii.) That the surface which he believes to be the most contractile is on the conver side at the bending-point. (iii.) The strong argument that the cells, when isolated by nitric acid and chlorate of potassium, exhibit precisely the same power of independent torsion as those of the Stipa- awn. The mechanism of the knee is, I believe, the same in Avena as in Stipa. Professor Hildebrand figures a section of this awn (which is morphologically identical with that of Stipa) as having two cavities answering to the hollow ribs of Stipa, and, like them, filled with cells containing chlorophyl in the young state. In this condition the ribs make an “The following trick is made use of when any thing in a family is lost, to make the person suspected confess the fault and restore what is lost. Take a beard of wild oat while 'tis greenish, and twist it in the shape of a little cross, giving it to the person suspected, and whose guilt they are pretty well assured of. Give also to the rest of the family little crosses, but made of different stalks, as hay and wheat ; put all these in а cut apple, and the little wild oat will 8row sensible of the moisture, untwisting itself and turning, to the great amazement of all the spectators.” 164 DR. FRANCIS DARWIN ON THE MECHANISM BY WHICH elongated turn in the direction opposed to the hygroscopic torsion of the awn; and this elongated “ reverse twist" is seen іп the mature awn when untwisted by wetting. This reverse twist is the consequence of the approach of the ribs to each other as they pass from the lower part of the awn to the point of bending; and I think that, as in Stipa, this alteration in form is connected with the longitudinal partition of the awn into a more and a less contractile half. HETEROPOGON MELANOCARPUS. The general appearance of the awn of this species is shown in fig. 2. In an ordinary room in winter it has but one knee, as shown in the figure ; but I have since found that by drying it at a higher temperature a lower knee appears; in the wet state a “ reverse twist" сап be made out. The seed has a sharp oblique point, and а well-developed plume of hairs; the awn, like that of Stipa, Avena, and Anemone, is twisted into a right- handed screw. There are no feathery hairs on the non-twisted part of the awn; but it is covered throughout its entire length with minute reflexed barbs. I find it capable of burying itself on becoming wet; but whether it does so on becoming dry I am not sure. HETEROPOGON CONTORTUS has a doubly bent twisted awn, of which the vertical part is roughly 13 mm., the part between the knees 10 mm., and the non-twisted part 4 cm. in length. It twists in the same direction as Stipa &c. ; and, when wetted, I have found it able to thrust itself into sand. ANDROSCEPIA ARUNDINACEA. The strongly bent and twisted awn is shown in fig. 3. I do not possess the complete fruit of this species, and could not make any trial of its burying-powers. ANTHESTERIA CILIATA somewhat resembles Heteropogon melanocarpus, with a bent and twisted awn, and 4 fine plume of hairs. Lagurus ovatus has a bent and twisted hygroscopic awn. And I find the following kinds of grasses mentioned as having bent and twisted awns— Aira, Arrhenatherum, Holcus*, Streblocheta nutans, Danthonia, Chetobromus Dregeanus, Macr ochloa arenariat. I think it is probable that these have the power of burying themselves. RANUNCULACES. ANEMONE MONTANA. Fig. 6 gives the general appearance and proportions of the achene of this plant. The slight hairiness of the awn probably aids in the distribution of the seed. It is exceedingly remarkable that the general features of the above-mentioned burying-seeds, all members of the order of Grasses, should be repeated in a Ranunculaceous plant. * ‘British Flora,’ Hooker. + Max Wichura, translated in Taylor's Scientific Memoirs, May and Aug. 1853. CERTAIN SEEDS BURY THEMSELVES ІМ THE GROUND. 165 We have :—(1) seed more or less pointed, with a plume of hairs directed backwards (see fig. 8) ; (2) vertical awn twisting, when dried, into a right-handed screw ; (3) hori- zontal, non-twisting portion rotating, when wet, in the direction of the hands of a watch, and coming into the same straight line with the lower part of the awn when the untwist- ing is complete; (4) cells of the twisting part capable of independent torsion. With some difficulty I succeeded in observing the seed of Anemone montana bury itself almost completely during the process of becoming wet. I do not know whether it is сара- Ме of being buried deeper as it dries. Le Maout and Decaisne * figure the achene of Anemone pulsatilla as bent and twisted. Max Wichurat describes Clematis azurea as having the appendages to the carpels twisted. : GERANIACEA. I merely mention this family to point out the general resemblance presented by the coccus and awn in some of its members to the burying-mechanisms already described. Fig. 5 (a dry Pelargonium-coccus) shows the flat, ribbon-like awn twisted into a right- handed screw, the knee, the horizontal non-twisted portion, the pointed fruit, and the plume of grappling-hairs. It is undeniably true, as Hanstein points out, that the external surface contracts more strongly than the internal one in drying; but this does not account for the direction of torsion ; Г find that, as in the above-described awns, the cells are capable of independent torsion. | I have made no experiments on the burying-powers of the Сегапіасеғ, as these һауе been already described by Hanstein, Roux, and Asa Gray. When we find among organisms belonging to widely different groups a curious struc- tural mechanism, repeating itself and performing in each case the same function, we conclude that this function is an important one in the economy of the plant. Thus, for instance, the importance of the distribution of seeds is pointed out by the existence of burs in widely different orders, such as the Composite (Burdock), Rosaceze ( Geum), and Rubiaceæ (Galium aparine)—of plumes to enable the seed to fly on the wind in many Compositze, Аросупасеге, Onagraceze, &c. bs Now in the burying seeds which we have been considering we have a similar case, а given function performed by essentially similar mechanism in plants belonging to widely different orders; and we accordingly conclude that the power of thrusting themselves into the ground is of special service to the seeds under consideration. "Two theories suggest themselves. i 22 (i.) My father has observed that certain seeds are almost incapable of germinating in the light, whereas they do soreadily in the dark. This fact suggested that the power of self-burial has been developed to remedy the injury which the incapacity of the seeds to germinate in the light must cause to the species. I therefore made a comparative trial to determine this point. A number of Stipa-seeds were placed in a vessel half filled with ‘lor’s Scientifie Memoi 853, p. 302. * English Translation, р. 175. + Taylor's Scientific Memoirs, Aug. 1853, р. 30 166 DR. FRANCIS DARWIN ON THE MECHANISM ВУ WHICH damp sand, and covered with а glass plate; half of the seeds were thrust into the sand* ; the rest were allowed to remain on the surface. Contrary to my expectation, the seeds exposed to the light began to germinate first. I regret that I did not make a second trial with another lot of seeds; I think, however, that the above result is sufficient to overturn the theory in question, at least as far as Stipa is concerned t. (ii.) The only other hypothesis which I have been able to form is, that by burying themselves in the ground the seeds are enabled to escape being eaten by birds. The fact that many of them are Graminez, and therefore likely to be sought after as food, favours this view 1. | The developmental stages through which any structural mechanism has passed is alwaysa most interesting question. Unfortunately I am as yet unable to enter into this question with respect to the burying seeds. We have seen that the mainspring of the mechanism is hygroscopie torsion; now many cells become twisted on drying, such as cotton-wool and bast fibres, the tubes of Zrinewn $, and curiously enough the beautifully striated cells which support the glands in Byblis gigantea. Again, hygroscopic torsion depends оп the spiral striation of the walls of the twisting cells. Now lamination or differentiation into more or less watery layers is probably a universal condition of the formation of cell-walls ; so that the materials, as it were, for the development of hygroscopic torsion are certainly existent. The variability in the torsion of the cotton-wool cell shows that a want of unanimity in the direction of torsion would be one of the difficulties to be overcome іп the process of development. Through what steps the hygroscopic sensitiveness has passed in development I am at a loss to say. Finally, I venture to hope that my explanation of the torsion of awns may ultimately throw some light оп other forms of torsion; and to this point I hope soon to direct my attention. % For whatever reason the seed is buried, 16 would seem desirable that it should remain undisturbed. Тһе joint by which the awn is attached to the seed seems adapted for this purpose ; for after remaining a few days in damp soil the awn сап be detached by the slightest touch. Roux points out the same thing in Erodium. + G. Roux, loc. cit. states that Erodium-seeds germinate better when buried to the “ normal” depth than when more or less deeply covered by soil. The details are not given in a manner to carry conviction. + The seed of Anemone montana (one of the burying seeds) contains a large quantity of endosperm ; the seed of this species is moreover, as far as I can make out, destitute of the acrid principle so common in the Ranunculacee, and which is found at least in some representatives of the neighbouring genus Clematis. Is it possible that the acrid principle may serve to prevent animals or birds eating the seeds of the Вапиисшасе» ? as we know to be the case with the Bitter Almond (* Variation of Animals and Plants,’ 2nd edit, vol. ii. р. 218). If so, the deficiency in the seed of Anemone montana of an acrid principle may be connected with the development of the burying-mechanism. $ Sachs, * Physiologie Végétale,’ (translation: Geneva, 1868) р. 453, quoted from Nügeli and Cramer, ‹ Pflanzen- phys. Untersuch.’ CERTAIN SEEDS BURY THEMSELVES IN THE GROUND. 167 DESCRIPTION OF PLATE XXIII. . Fruit and awn of Stipa pennata (natural size). p, the sharp point; v, the vertical twisted part of the амп; Ё, and k, the lower and upper knees; f, the non-twisted and more or less hori- zontal part of the awn. Fruit and awn of Heteropogon melanocarpus. Letters corresponding to fig. 1. . Awn of Androscepia arundinacea. Letters as above. . Fruit and awn of Avena elatior. x 6. . Carpel and beak of Pelargonium. Letters as above. x 31. . Achene and tail of Anemone montana. Letters as above: natural size. . Lower part of the fruit of Stipa pennata. х 6$. . Achene of Anemone montana. х 5 (about). Upper part of v in fig. 1. х 7 (about). r and r аге spirally running ribs, seen in section at r and r in figs. 11, 12, 13. 10. Cotton-wool cells. а, in dilute, 6, in stronger sulphuric acid (from Nägeli). х 200. 11. Section of awn of Stipa pennata close to seed. т and т hollow spiral ribs (see fig. 9); с, central | vascular bundle; m and m smaller masses of the same tissue; с, т, and m are non-hygroscopic. | х 80. 12. Section of awn of Stipa pennata at k,, fig. 1. Letters as in fig. 11. х 80. 13. Section of awn of Stipa pennata at k, fig. 1. Letters as in figs. 11 and 12. гу, line dividing awn into a more and a less hygroscopic part. On the other dotted line see text, p. 160. x 80. 14. External cells from the awn of Stipa pennata, isolated with nitric acid and chlorate of potash and then dried, to show the power of independent torsion residing in the cells of the awn. x about 100. A os = t0 ON ф OW O 0 SECOND SERIES.—BOTANY, VOL. I. 2A A R Darwin del. Trans Linn бос.бен 2 Вот Voll Tas.23 % d E ААА SAGE BU Sx I о ШО OSIN УСК Tae КЕ БӨ IOS 208% ре RASS ES “л a SAIH РРА РРР 2 КЕ; N Ў N `A 5 RN : x NY C.Berjeau lith. TORSION SEEDS, AWNS & SECTIONS OF GRASSES, кс. ТНЕ TRANSACTIONS OF ІНЕ LINNEAN SOCIETY OF LONDON. SECOND SERIES.—BOTANY VOLUME I. PART THE FOURTH. MISSOURI BOTANICAL GARDEN. sae oO NS PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET : SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. M.DCCC.LXXVI. CONTENTS. PART IV.—DECEMBER, 1876. X. On the African Species of the Genus Coffea, Linn. Ву W.P. Наввх, MA., к у v. d. ХТ. On the Origin of Floral ZEstivations. With Notes on the Structure of the Cruci- Jerous Flower, on that of Adoxa, and on the Corolla of Primula. By the Rev. ЧЫ Ийи ША FLS . a x o4 ош л XII. On some points in the Histology of certain Species of Corallinacee. Ву Major- General R. S. Netson, R.E., and Prof. Duncan (М.В. Lond.), Pres. Geol. Soc. Communicated бу Dr. Маша PLS. о... м ХПІ. On the Minute Structure and Mode of Growth of Ballia callitricha. Ву W. Авснев, F.R.S., М.В.Г.А. Communicated by W. Т. THISELTON DYER, NEUE BRNO 22. 22277... ‚л [ 169 1 X. On the African Species of the Genus Coffea, Linn. Ву W. Р. Нікі MA, PLS. (Plate XXIV.) Read April 20th, 1876. AS at present understood, the Linnean genus Coffea belongs to the Old World ; and the numerous American species that have been previously referred to it, now find places in other genera. Allthe species most valuable for economie or commercial purposes are confined to Africa or are of African origin. Among the Indian species, at least one, C. bengalensis, Roxb., a native of Silhet, was in former times cultivated in Bengal for the growth of coffee ; but being far inferior and not productive, it has been discarded on the introduction of the African plants. Some of the Indian species originally deseribed as members of this genus, belong to the following different genera :—Psychotria, Linn. ; Prismatomeris, Thw. (C. tetrandra, Roxb.); and Pavetta, Linn. The American species belong to Appunia, Hook. f. (C. tenuiflora, Benth.), Coussarea, Aubl, Furamea, Aubl., Psychotria, Linn., Palicourea, Aubl., and Rudgea, Salisb. In Africa, besides the true species of Coffea, other berries are employed as coffee; but they are not of commercial importance : an example of this occurs in Feretia apodanthera, А. Rich. Randia genipeflora, DC., is called wild coffee by the people at Sierra Leone and in the island of Fernando Po, according to Barter's notes in the Kew Museum and Herbarium. Of the species included in the present paper, in all fifteen, seven have been already published with descriptions, and the remaining eight are new or have not been pre- viously described. One species, C. arabica, the original species of the genus, is believed to be indigenous in Abyssinia, also on the shores of the Victoria Nyanza and in the district of Golungo Alto in Angola, doubtfully also along the Mozambique coast, while the white-berried variety comes from Sierra Leone. Another species, C. liberica, belongs to Liberia, to Golungo Alto, and to Sierra Leone, though perhaps only cultivated in the latter place; two other species, C. stenophylla and C. Afzelii, are peculiar to Sierra Leone ; one, С. microcarpa, to Senegambia; another, С. rupestris, to Abbeokuta ; another, C. brevipes, to the Cameroons Mountains; another, C. subcordata, to Old Calabar ; one, С. jasminoides, is common to Old Calabar, to the Niger region, and to Golungo Alto; another, C. melanocarpa, is peculiar to Golungo Alto; another, C. hypo- glauca, to the district Pungo Andongo of Angola; two, С. Zanguebarie and С. racemosa, to the Mozambique district ; and, lastly, two, C. mauritiana and С. macrocarpa, to the Mascarene Islands of Bourbon and Mauritius. In other words, Upper Guinea contains 9 species, Nile Land 1, Lower Guinea 5, and Mozambique district with the Mascarene Islands 4 or 5. SECOND SERIES.— BOTANY, VOL. I. 2B 170 MR. W. Р. НЕВМ ON THE AFRICAN SPECIES The following key determines the classifieation of the species :— Coffeæ speciebus Africanis clavis adaptata. Glabre, plerzeque sempervirentes. Flores axillares. Calycis limbus brevissimus, annularis vel denticulatus. Antherz omnino exsertæ. Bracteole obtuse vel apiculate, calyce breviores. Flores glomerati, rarius solitarii. Corolla pentamera ее 1. Corolla hexamera vel heptamera vel octamera. Bacce teretes. Folia subcoriacea. Flores glomerati. Folia breviter acuminata, 6-12-polhcaria . . . . 2. Flores terni vel gemini vel solitarii. Folia caudato-acuminata, 25-5- Нн iE л: 3 Васса nervis longitudinalibus angulate. - Копа chartacea .. . . = ..4 Bracteole per paria lanceolate acut: et deltoideo-apiculate, calyce lon- BENE DAMEN S oS gr zs... Anthere semiinclusee. Бассғе utrinque rotundate. Folia chartacea . . . . . . . . . 6. Bacce utrinque acute. Folia coriacea. Folia 2-3-pollicaria, tenuiter coriacea. Corolla j-polliearis . . . . 7. Folia 5—6-роШсагіа, valde coriacea. Corolla j-pollicaris. . . . . 8. Calycis limbus latus, lobis 5 rotundatis. Stipulz lanceolatz subulatz. Bracteole parve, non foliacee. Bacce subglobose, solitariw vel subgemine. Folia 3-9 poll. longa, Oe 2222-22-20 Bacce ellipsoidex. Flores secus ramulos ad axillas fasciculati. Folia 24 pou: longa, 6-7 poll. lata. . 0; , рю Stipulz e basi lata ovata apiculate. Bracteole foliacee . . . . . . . .l. Non omnino glabre. Folia decidua. Flores terminales axillaresque vel laterales vel ramulos breves laterales terminantes. Folia ovalia, puberula vel glabra. Antherz omnino exsertz. Flores cum foliis coevi . . . . . . . . . .12. Antherz fere incluse. Flores ргаесосез. Bracteole herbacez. Calycis limbus truncatus vel denticulatus. Corolla 9... Co oe ; ТЕ. Bracteole glumacee. Calycis limbus rotundate lobatus. Corolla sæpius 7—6-loba 1 . . . . . . . . . . . . . . . . . . 14. Folia ovato-lanceolata, tuberculis plurimis consita, scabra . . . . . . . ‚15. 1. COFFEA ARABICA, Linn. Sp. РІ. p. 172 (1753). C. laurifolia, Salisb. Prodr. Stirp. Hort. Chapel Allert. p. 62. п. 1 (1796), non Kunth. ARABICA. LIBERICA. . STENOPHYLLA. ZANGUEBARLE. BREVIPES. MELANOCARPA. MAURITIANA. MACROCARPA. HYPOGLAUCA. MICROCARPA. AFZELII. SUBCORDATA. RUPESTRIS. JASMINOIDES. RACEMOSA. Hab. Indigenous in Abyssinia, also on the slopes above the Victoria Nyanza ( Grant), and in Angola ( Welwitsch), and probably in some of the intervening countries ; cul- tivated and often occurring spontaneously in Arabia and nearly all tropical and sub- tropical countries, including Madagascar and Natal; said to be wild along the Mozam- bique coast [see Klotzsch in Peters, Mossamb. р. 291 (1862)]. It, with its numerous cultivated varieties, is the source of most of the coffee of commerce, OF THE GENUS COFFEA, LINN. 171 Dr. Welwitsch found this species quite indigenous in primitive woods throughout nearly the whole district of Golungo Alto, as well as frequently cultivated ; he describes it as a small tree, 10—15 feet high, with the trunk 9-12 inches in diameter at the base, and with horizontal or even nodding branches, which in old age become unilateral; it belongs to the * Regio montoso-sylvatica’ of Welwitsch, which ranges from 8° to 10^ S. lat. and from 1000 to 2400 feet of altitude. The commercial varieties of common coffee depend mainly upon their shape, size, and colour. The shape of the berries and of the seeds is said to depend on the particular portion of the plant on which they grow, those near the extremities being often obliquely rounded by the abortion of one of the seeds in the fruit; the size and succulence depend on the nature of the locality—those berries grown in a dry, hot, and rocky climate being smaller, less fleshy, and of superior flavour; and the colour depends on the degree of maturity attained by the fruit at the time of gathering. Some or all of these differences may also partly depend on the variety of the plant employed; but it is not known for certain whether the Mocha coffee is a distinct variety, or, as seems more probable, only due to a particular treatment and cultivation. The following variety I have described, since it may prove to be a distinct species, characterized by white berries бе. ; the flowers are unknown. Var. leucocarpa, fruticosa glabra, foliis ellipticis vel ovali-oblongis breviter obtuse acuminatis basi cuneatis tenuiter coriaceis, venis lateralibus inconspicuis sub-6—7-jugis eglandulosis, stipulis e basi late ovata connata apiculatis petiolum brevem sub:equanti- bus, floribus axillaribus subglomeratis, bracteolis ovatis quam pedicelli breves fruc- tiferi brevioribus, baccis albis pisiformibus solitariis vel geminis, calycis limbo fructiferi obsoleto. Hab. ad Sierra Leone, legit mense Junio anni 1841 Th. Vogel! no. 174. Folia 3-6 poll. longa, 1-2 poll. lata. Ап species distincta ? 2. COFFEA LIBERICA, sp. nova; Hort. Bull. Pl. XXIV. C. fruticosa vel arborea glabra lucida sempervirens, ramis horizontaliter patentibus, foliis elliptico-obovatis oblongisve breviter acuminatis basi euneatis vel obtuse angustatis subundatis tenuiter coriaceis 1—1-pedalibus, nervis lateralibus utrinsecus 8-10 in alis puncto secretorio instructis, stipulis late ovatis apiculatis basi connatis quam petioli sub- brevioribus, floribus axillaribus glomeratis subsessilibus pollicaribus, bracteolis connatis calyculatis depresso-deltoideis vel obtusis plerisque subtruncatis calyce brevioribus, calycis limbo annulari brevissimo, corollæ lobis 7-6 ovalibus obtusis tubum subzequanti- bus, antheris 7-6 omnino exsertis semipollicaribus, filamentis à-pollicaribus, stylo exserto bifido, baccis ellipsoideis -pollicaribus vel ultra, seminibus semipollicaribus vel ultra. C. arabica, Benth. in Hook. Niger Fl. p. 413 (1849), part., non Linn. Hab. ad Sierra Leone, legerunt Afzelius! et Daniell! (cult.); in Monrovia, testibus Th. Vogel et Daniell; etiam in Angola (Golungo Alto et Cazengo) legit Welwitsch ! The source ofthe Liberian coffee, and probably also of the Cape-Coast coffee; it is said to be far superior to C. arabica, Linn., having larger berries, of a finer flavour, and being at the same time more robust and productive. 282 B 172 MR. W. P. HIERN ON THE AFRICAN SPECIES The name Coffea liberica was given by Mr. W. Bull, F.L.S., of Chelsea, in order to distinguish from the common coffee of commerce young plants which he obtained from seeds of Liberian coffee that had been sent from the west coast of Africa. Не has since distributed the plant extensively under this name; but he had no idea whether it would prove to bea good species from a botanical point of view, nor did he intend the name to have any such sense or indication. However, as it turns out that the plant really constitutes a new species and requires a botanical name, I have thought it more con- venient to adopt Mr. Bull’s horticultural name, which has already received a wide distribution amongst plant-growers, rather than to take a manuscript name which Afzelius long ago gave to the same species in his herbarium, but which has never been published, and is entirely unknown to the publie. Mr. Bull's name first appeared in his * Retail List of new beautiful and rare Plants, No. 97, page 4 (1874), and has been repeated in all the subsequent editions; he has given no botanical description. The plant appears to have been first grown in this country in the year 1872 at Kew Gardens, from seeds contributed from Cape Coast, but only in very small quantity ; in the following year living plants were imported from the west coast of Africa by Mr. Bull, who also in the same year obtained plants from seed; and at the same time numerous seedlings were raised at Kew. Itis now produced on a large scale in this country for distribution to the various coffee-growing parts of the world. 3. CorrEA sTENOPHYLLA, G. Don, Gen. Syst. iii. p. 581, n. 4 (1834). C. arabica, Benth. in Hook. Niger Fl. p. 413 (1849), part., non Linn. Hab. Sierra Leone, Afzelius ! and others. In the Appendix to the Report of the Sierra-Leone Company, published in the year 1794, Afzelius states, on page 173, that “ Coffee-trees are found of two distinct species, both nondescript; but whether of any use is not yet ascertained." This remark applies to this and the previous species. He subsequently discovered the economie value of at least this species. According to G. Don, the seeds of this species are roasted and used as the common coffee, and are even considered superior to it. Afzelius, as has been explained, had recognized this as a distinct species ; and among his manuscripts there remains à long description of it under an unpublished name of his own. It is the Highland coffee of Sierra Leone, according to Dr. Daniell. 4. COFFEA ZANGUEBARLE, Lour. Fl. Cochinch. p. 145, п. 3 (1790). Amazona africana, Spreng. Syst. Veg. ii. p. 126 (1825). Hab. Zanzibar coast, Loureiro; Mozambique, Forbes!, Loureiro (cult.). The seeds are said to be used in the same manner as those of C. arabica, Linn. 5. COFFEA BnEVIPEs, Hiern, sp. nova. C. fruticosa glabra nitida, ramulis gracilibus, foliis elliptico-obovatis acuminatis basi valde cuneatis firmiter chartaceis, venis lateralibus utrinsecus 6-8 gracilibus, stipulis e basi ovata valde vel rigide apiculatis petiolum brevem excedentibus, floribus axil- ОЕ THE GENUS СОҒҒЕА, LINN. 173 laribus solitariis subsessilibus eum foliis coævis, bracteolis per paria lanceolato-acutis et deltoideo-apiculatis calycis limbum obsoletum excedentibus, coroll, lobis 5-6 ovalibus vel oblongis, antheris 5-6 omnino exsertis filamento bis longioribus, stylo exserto bifido, baccis lzevibus ellipsoideis semipollicaribus solitariis. Hab. in montibus Cameroons ad 2000-3000 ped. alt., legit mense Decembri anni 1862 G. Mann !, no. 2158. Frutex 4-6-pedalis. Folia 4—85 poll. longa, 13-2} poll. lata. Flores dum expansi }+} poll. longi. 6. COFFEA MELANOCARPA, Welw. MS. in Herb. С. fruticosa vel arborea glabra, ramulis erecto-patentibus, foliis ellipticis acuminatis basi cuneatis breviter petiolatis stipulos parvos obtusos excedentibus chartaceis, venis lateralibus utrinsecus 8—5 gracilibus, floribus pentameris sesquipollicaribus subsessilibus axillaribus subterminalibusque solitariis subgeminisve cum foliis covevis, calyce basi bracteolis omnino connatis calyculatis cincto, limbo annulari minimo, сого] lobis ovalibus i-i-polliearibus patentibus vel subreflexis, antheris majore parte inclusis subsessilibus stylo incluso, baccis demum aterrimis ellipsoideis vel subdidymis $—j- pollicaribus. Hab. in sylvis clarioribus circiter Sobati Bango et Bumba ad fluv. Delambra, in territorio Angolensi Golungo Alto non infrequens; mensibus Novembri et Decembri anni 1855 legit Welwitsch ! Frutex 4—6-pedalis, hine inde arbuscula ejusdem altitudinis. Folia 14-6 poll. longa, 5-24 poll. lata. Gemmze florum exsudatione tenui vitrea fragili tecta. 7. COFFEA MAURITIANA, Lam. Encycl. Méth. i. р. 550, n. 2 (1783); Illustr. ii. p. 238. n. 3109, t. 160. f. 2 (1800). C. arabica В, Willd. Spec. Pl. i. p. 974. n. 1, var. (1797), non Linn. C sylvestris, Willd. in Roem. et Schult. Syst. Veg. v. p. 201. n. 28 (1819), fide DC. Prodr, iv. 499 (1830). Hab. Bourbon, Richard !; Mauritius, J. Grey! In high mountainous forests. Called Café-marron in Bourbon. It is mixed with other kinds of coffee ; taken alone, it is said to have intoxicating properties. 8. COFFEA macrocarpa, A. Rich. іп Mém. Soc. Hist. Nat. Par. у. p. 168 (1834). „Наб. Mauritius, Bouton! Іп dense mountainous forests. 9. Correa HYPOGLAUCA, Welw. MS. in Herb. Arborea glabra, ramis patentibus, ramulis patulis, foliis ellipticis acuminatis basi plus minusve cuneatis tenuiter coriaceis nitentibus subtus glaucescentibus vel albescentibus, venis lateralibus utrinsecus 5-6 gracilibus, stipulis e basi lata lanceolato-subulatis acutissimis caducis petiolum excedentibus, baccis solitariis vel subgeminis axillaribus Subsessilibus vel breviter pedicellatis subglobosis basi bracteolis ciliolatis imbricatis . parvis rotundatis suffultis (immaturis viridibus pisi mole) apice calycis lobis 5 persisten- tibus imbricatis bracteolis conformibus coronatis. 174 MR. W. P. HIERN ON THE AFRICAN SPECIES Hab. іп sylvaticis prope Catete in territorio Angolensi Pungo Andongo, cum fruc tibus mensibus Decembri anni 1856 usque ad Maium anni 1857 legit Welwitsch ! Arbuseula 8-15-pedalis et altior vel arbor parva 15-20-pedalis, ut videtur semper- virens, coma laxa, trunco recto gracili; Folia 3-9 poll. longa, 14-84 poll. lata, petiolo i-i-pollieari ; stipulis $—1-pollicaribus. 10. COFFEA ? MICROCARPA, DC. Prodr. iv. p. 499, n. 4 (1830), non Ruiz et Pavon. С. Perrottetii (?), Steud. ex Buek, Index DC., pars 1, praef. p. ix (1842). Hab. Senegambia, Leprieur & Perrottet. I have seen no specimen of this species. 11. COFFEA Ағ?Етлі, Hiern, sp. nova. C. glabra, ramulis patulis rubescentibus, foliis ellipticis vel ovato-ovalibus obtuse acuminatis basi cuneatis chartaceis, venis lateralibus utrinsecus 4—5 gracillimis, stipulis e basi ovata apiculatis petiolum brevem sub:equantibus, floribus pentameris geminis axillaribus pedunculatis basi bracteolatis, pedunculis prope basim bracteis 4 parvis oppo- sitis preeditis, bracteolis foliaceis, calycis limbo 5-dentato, corollæ lobis 5 ovalibus ob- tusis, antheris 5 sessilibus faucibus nudis infra os insertis, stylo glabro. Hab. ad Sierra Leone, legit Afzelius ! Folia 14-3 poll. longa, $—11 poll. lata. l have seen only a single specimen of this species, in Afzelius's herbarium, in very young flower. 12. COFFEA suBCORDATA, Hiern, sp. nova. C. fruticosa, ramulis teretibus hispidulis patentibus gracilibus apice hispido-pubescen- tibus, foliis ovalibus obtuse acuminatis mucronulatis basi subcordatis firmiter chartaceis supra costa excepta glabrescentibus infra secus costam et nervos laterales 4—5-jugos hispidulis, stipulis e basi deltoidea caudatis setoso-apiculatis deciduis petiolum brevem sequantibus, floribus vel terminalibus glomeratis et axillaribus solitariis vel geminis et ramulos abbreviatos laterales terminantibus (4-) 5—6 (—7)-meris fragrantibus cum foliis coævis, bracteolis calycem excedentibus herbaceis quibusdam subfoliaceis, calycis limbo glabro 5-fido lobis rotundatis, corolle nives tubo gracili lobis ovalibus obtusis, antheris omnino exsertis, stylo exserto glabro bifido, baccis rubris. Hab. ad oppidum Old Calabar frequens; legit mense Martio anni 1863, cum floribus, Rev. W. C. Thomson ! no. 35. Folia usque ad 31 poll. longa et 13 poll. lata pervenientia. Flores $—-pollicares, suaveolentes. 13. CorrEA RUPESTRIS, Hiern, sp. nova. с. fruticosa Subglabra, ramis subteretibus cinereis patulis ramulis puberulis, foliis deciduis ovalibus utrinque angustatis glabris chartaceis, stipulis ovatis acutis vel apicu- | ОЕ THE GENUS COFFEA, LINN. 175 latis, floribus przecocibus pentameris candidis sessilibus ad ramulorum apices glomeratis fragrantibus, bracteolis herbaceis late ovatis apiculatis calycis limbum subtruncatum vel denticulatum brevem excedentibus, corolle lobis ovalibus obtusis tubum zquantibus, antheris sessilibus majore parte inclusis, stylo bifido. Наб. in rupibus ad oppidum Abbeokuta, legit Barter ! no. 3343. Frutex parvus; folia (juvenilia) pollicaria; flores pollicares. 14. COFFEA JASMINOIDES, Welw. MS. in Herb. C. fruticosa plus minusve sarmentosa ramosissima, ramis patentibus decussatis pube- rulis vel glabratis, foliis ovalibus acuminatis rarius apice rotundatis ssepe mucronulatis basi rotundatis vel rarius subcuneatis oppositis vel rarius quasi quaternate verticellatis deciduis chartaceis nervis exceptis glabratis supra nitidis infra secus costam et 4-5-jugos nervos puberulis vel brevipubescentibus breviter petiolatis, stipulis e basi brevi lata subtruncata setaceo-apiculatis petiolum pubescentem subexcedentibus, floribus candidis extus заре roseo-sanguineis vel niveis fragrantibus sepe subpendulis solitariis vel sub- geminis lateralibus vel ramulos abbreviatos laterales terminantibus preecocibus subsessi- libus basi bracteolatis, bracteolis glumaceis ovatis vel lanceolatis acutis imbricatis caly - cem excedentibus rarius quibusdam subfoliaceis, calycis urceolati limbo ssepius 8-fido lobis rotundatis rigidulo-membranaceis glabris, corolla 7—6- rarius 5-loba caduca, antheris sepius 7-6 erectis subinclusis filamentis brevissimis, stylo bifido lobis patulis crassis lanceolato-spathulatis obtusis intus stigmatosis. Hab. in sylvis et dumetis densis ad basim et declivia montium Serra do Alto Queta territorii Angolensis, Golungo Alto legit Welwitsch ! no. 2572, Decembr. 1854 et Octobr. 1855 с. flor., etiam Novembr. 1855 et Junio 1856 c. fr.; in Nigritia ad Onitsha legit Barter ! no. 1249 (suffrutex subscandens, baccis albis); etiam ad oppidum Old Calabar legit Rev. W. С. Thomson ! по. 37, freq. Frutex 2-5-pedalis; folia 11-3 poll. longa, 3-14 poll. lata. Flores pollicares vel sesquipollicares ; расса 3-3 poll. diam. 15. Correa ВАСЕМОЗА, Lour. Fl. Cochinch. p. 145, n. 2 (1790), non Ruiz et Рахов (1799). C. ramosa, Roem. et Schult. Syst. Veg. i. р. 198. n. 10 (1819). C. mozambicana, DC. Prodr. iv. p. 500. n. 18 (1830). Hab. Mozambique Island, Loureiro. The seeds are said to be used for coffee. I have seen no specimen. Excluded Species. CoFFEA LAURINA, Poir. in Encycl. Méth. Suppl. ii. p. 14. n. 15 (1811), is Craterispermum laurinum, Benth. Sierra Leone. | | Correa сүмозА, Willd. ex Roem. et Schult. Syst. Veg. у. p. 201. n. 27 (1819), is doubt- fully referred to Chasalia Fontanesii, DC. Bourbon. 176 MR. W.. P. HIERN ON THE AFRICAN COFFEZ, COFFEA CAPITATA, Sieb. Fl. Maurit. Ехѕісе. n. 335, is Chasalia coffeoides, DO. Mau- ritius. COFFEA DIVARICATA, Tausch in Sieb. Fl. Maurit. Exsicc. ii. n. 271, is Chasalia divaricata, DC. Mauritius. COFFEA HIRSUTUS, б. Don, Gen. Syst. Gard. & Bot. iii. p. 581. n. 5 (1834), is Cremaspora africana, Benth. Sierra Leone. CorrEA KRAUSSIANA, Hochst. in Flora, 1842, p. 237, note, is Kraussia floribunda, Harv. Natal. DESCRIPTION OF PLATE XXIV. Plate of Coffea liberica, from a specimen in the British Museum, collected by Dr. Daniell in January 1856, at Sierra Leone, where it was cultivated : the flowers have been preserved in liquid. Тһе principal figure shows a flowering branch (natural size), with a small lump of clear gummy secretion at the apex, with flower-beds in the next axils, with flowers (some fully expanded) in the next, and with very young fruits on the lowest axils that are included in the figure. Fig. а shows a cluster of three calyces, also natural size, with the calyculate upper bracteoles covering the short pedicels, the middle calyx shown after the fall of its corolla, the right-hand one with a fully expanded corolla, and the left-hand one with a young unexpanded corolla. Fig. b shows a vertical section, somewhat enlarged, of the ovary and epigynous disk by a central plane perpendieular to the septum ; the solitary ovules are seen to be attached near the middle. Fig. c shows a pyrene as seen from the face, natural size. Fig. d shows a lateral view of the same, also natural size. Fig. e shows a dorsal view of a seed (natural size) cut away rather obliquely towards the base so as to expose the embryo with its inferior radicle. [177 J XI. On the Origin of Floral ZEstivations. With Notes on the Structure of the Cruciferous— Flower, on that of Adoxa, and on the Corolla of Primula. Ву the Rev. G. Henstow, M.A., F.L.S., F.G.S. | (Plate XXV.) Read June 1st, 1876. INTRODUCTION. THE idea of endeavouring to trace out any laws that may exist amongst floral æsti- vations arose in my mind from observing the large number of diagrams given in Le Maout and Decaisne’s * Descriptive and Analytical Botany, in which the æstivations of about 250 genera of Dicotyledons and of 50 of Monocotyledons are diagrammatically expressed. I therefore proceeded to draw up an enumeration of all the different kinds, and to ascertain what percentage each had amongst the genera, and what connexions there might be amongst those different kinds. I then examined a number of living flowers of several genera, but soon found my work had been to some extent inefficient as based upon very insufficient data *. So far from each diagram given being the sole example or even always typical of a genus, I found it to be the commonest thing for a species to have a very considerable amount of variation, as will be seen at the close of this paper. As an example, the diagram of Primula, p. 529 (Engl. ed.) 4, is not the only kind, as this genus can boast of at least eight varieties in the Primrose alone. Again, the diagram of the male flower of Hydrocharis, p. 154, does not correspond with the figure of the flower given there, in that the inner whorl of the figure is “ tristichous," while it is represented as “ con- volute " in the diagram. Lastly, two diagrams of Hermannia are given on p. 284, as if it were an unusual occurrence for a flower to have its whorls running spirally in opposite direetions, whereas it is an extremely common occurrence. Hence, if any one should be led to imagine these diagrams to be in all cases characteristic or the sole kinds of æstivation peculiar to each genus respectively, he will be much mistaken. It may be observed, however, that nothing is said respecting them whatever, as to whether they do represent typical forms or not. Оп the other hand, the estivations of many genera are remarkably constant, such as the universally **convolute to the left” corolla of Myosotis; or again, the valvate calyx with contorted corolla of Mallow, though even here the “ imbrication"' may be either to the right or left; and when this is the case, it may become, as Robert Brown pointed out, a valuable diagnostic character. * The draught of this paper was made on recovering from a long illness, when I had but few opportunities of testing the conclusions given from the study of the diagrams alone. I, however, give these conclusions, because the reader will be able to refer to the diagrams in that excellent work, and because the deductions made from a comparison of those diagrams is generally very similar to that derived from the examination of the variations in the flower-buds of living plants. Sets t The references to pages throughout this paper always refer to Le Maout and Decaisne's work. SECOND SERIES.—BOTANY, VOL. I. 2c 178 THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. Little attention, indeed, appears to have been paid to :estivations until that eminent botanist called attention to their importance *. Since then nearly all text-books describe the principal forms, but without tracing any or but slight connexions between them. If we turn to our latest authorities, we find Le Maout and Decaisne, for example, devoting two pages (86, 87) to the subject in describing, as usual, the different kinds, and also attempting to explain the “ papilionaceous " and ** cochlear” by, as I believe, а wrong application of the “ quineunx ;" while in Sachs's * Text-book of Botany,’ “ sestiva- tion," “ imbricate,” * contorted,” and “ valvate," are words conspicuous by their absence from the index as well as from the text. Many diagrams occur in that work; but the subject of zestivation seems to have been totally ignored. The laws of phyllotaxis have to some extent been adapted to the floral whorls, more especially and obviously to the quincuncial or ? plan. The very frequent occurrence of this latter amongst sepals and petals is of itself a sufficient indication that phyllotactical laws are in certain cases applicable—as also the fact that the outer whorls of flowers are in some cases composed of many parts and spirally arranged, especially when there is no manifest “break” between the whorls, as in Calycanthus, Cactus, &c. Іп these cases the $, 3, 7s, and even зг arrangements may be detected. II. THE Forms оғ ÆSTIVATION, The following include those usually described, аз well as one or two in addition. 1. Distichous equitant, or the 4 arrangement in which one of the two opposite parts overlaps the edges of the other, as in Papaver (Tab. XXV. fig. 1). The half-equitant is a modification of this, in which one edge of each part overlaps one of the edges of the other part respectively, as in the calyx of Poppy and the corolla of Circea (Tab. XXV. fig. 2). 2. Tristichous or the X arrangement. This is especially characteristic of Mono- cotyledons. It may be observed that it is identical with the “ imbricate proper" when the whorl has only three parts, so that the first overlaps the second, and the latter overlaps the third, which last also underlies, with its opposite edge, the first (Tab. XXV. fig. 3). 3. Pentastichous, quincuncial or 2 arrangement, in which the parts of the whorl cor- respond with those of one projected ** cycle” of that plan (Tab. XXV. fig. 4). 4. Half-inbricate.—l apply this term to а large number of instances ranging from complete regularity to such extremely irregular forms as the corollas of the Violet, the Pea, and the Snapdragon. It is produced by a slight modification of the quincuncial— namely, by the edge of the second part being overlapped by instead of itself overlapping the fourth part (Tab. XXV. fig. 5). 5. Imbricate propert, in which the parts of а whorl overlap each other in succession, * А short summary of the history of the nomenclature of sestivations has lately appeared from the pen of Prof. А. Gray, іп the American Journal of Sciences and Arts, vol. x. Nov. 1875, and is reprinted in the Journal of Botany, new series, vol. v. p. 53. + І call this “ imbricate proper," as the word “ imbricate” is also used technically for an indefinite number of parts, as in the scales of Calycanthus &c., which overlap like tiles on a roof. Linnæus appears to have included all these forms (viz. 3, 4, 5, and 6) under this term (see Prof. Gray's paper 1. с.). THE REV. С. HENSLOW ОХ THE ORIGIN ОҒ FLORAL ASTIVATIONS. 179 so that one, the first, is entirely outside the adjacent parts, while the last is entirely within. This is deduced from the last by allowing the third member to slip one edge under the fifth, in addition to the second being under the fourth (Tab. XXV. fig. 6). б. Convolute and contorted.—In these each part of the whorl overlaps one adjacent part by one of its edges, and is in turn overlapped by the other adjacent part (Tab. XXV. fig. 7). Thisis deduced from the preceding in a similar manner to that—namely, by passing an edge of the first part under that of the third. Hence it will be seen that Nos. 3, 4, 5, and 6 are closely connected, and may be briefly expressed by saying the half-imbricate is formed from the quincuncial by having the 4th member over the 2nd, the imbricate proper from the half-imbricate, by having the 4th over the 2nd and the 5th over the 3rd, the convolute from the imbricate proper by having the 4th over the 2nd, the 5th over the 3rd, and the 3rd over the 1st *. 7. Valvate.—In this the edges of adjacent parts touch only, but do not overlap. The two main varieties, besides the simply valvate, where the edges just touch, are :— Induplicative, where the edges are rolled inwards; reduplicative, where the edges are rolled outwards. 8. Straight or open.—In this the parts of the whorl are not sufficiently developed to meet even as in valvate, and therefore the wstivation may be said to be indeterminate. ПТ. Tur DEGREES OF FREQUENCY OF THE VARIOUS KINDS ОҒ /ÉSTIVATION. The following percentages are deduced from Le Maout and Decaisne, as stated above. І. Dicotyledonous Angiosperms. А. IMBRICATIVE. 1. Equitant Оају зз уу ул або 4 p. c. x СОМА > -= ac 2.2. 5 2 2. Tristichous, Calyx, or Corolla . . . д ITI 3. Pentastichous, Calyx, with Cor. various 2 A Е Corolla, with Calyx ,, ds IU ees x both Calyx and Corolla 2” li 4. Half-imbricate (excluding Papiliona- ceous and cochleate varieties) . . » 15 4 5. Imbricate proper, Calyx, apparently very rare, as in Cyclamen, p. 529. Imbricate proper, Corolla .... » 9, 6. Convolute, Calyx, apparently rare. Dam . . i x s = 16 .„ 33 * Prof. A. Gray’s paper is chiefly directed to the settlement of the proper term for this kind; and һе сопсіпдев that “ eonvolute" is preferable to ** contorted," as the latter does not differentiate corollas with a torsion, the petals of which have each a spiral axis, from those with no torsion, or whose axisis erect. А third term, * obvolute," would, he observes, claim priority as having been used by Linnsus ; but it has not been in common usage. Ав the Professor concludes with the words, “ I should hope rather than immediately expect that this use would become general,” I have the satisfaction of fulfilling those expectations immediately ; for although this paper was written before Prof. Gray’s was published here, and I had adopted the word “ contorted,” I have altered it in every case to “ eonvolute," in accordance with his wish. 202 180 THE REV. б. HENSLOW ОМ THE ORIGIN OF FLORAL ASTIVATIONS. B. VALVATIVE. 7. Valvate (including varieties). Calyx valvate, with Corolla various. occurs about 11 p. c. Corolla valvate, with Calyx various. 5 > 8. Straight or open. Calyx open, with Corolla various 8 25 75 Corolla open, with Calyx various . a os II. Petaloid Monocotyledons. 1. Equitant is represented in 3 genera only besides Grasses. 2. Tristichous, outer perianth whorl . . ў 40 ,, Re inner perianth whorl . . 5 60 „ 8. Pentastichous occurs in Paris polyphylla, p. 859. 6. Convolute, represented in 4 genera only. КИ a S c 3 20-24 oc ш їп. аза e 5 80 «5, IV. REMARKS ON THE PRECEDING VARIETIES OF /ÉSTIVATION. 1. Equitant.—The proportion of flowers which have equitant whorls, either entirely, when one member of the pair is embraced by both the edges of the opposite member, or half-equitant, in which each of the pair overlaps one edge of the other, occurs in corollas about 12 р. с. and in calyces about 4 р. с., as calculated from the diagrams of Le Maout and Decaisne. Probably this is not far from being correct. With regard to the origin of this kind, I am inclined to think it is due to a reduction of parts from the tristichous or pentastichous arrangements, and in some cases may be a retention of a primitive condition. If one member of а tristichous whorl be suppressed, the other two will most probably become equitant, as flowers almost * invariably arrange their members either radially or bilaterally symmetrically. Such, I suspect, will account for this sstivation in poppies, for example, where the calyx is half-equitant and the corolla equitant (Tab. XXV. fig. 1). Papaver and its nearest allies of the tribe Eupapaveree are usually binary or quaternary in their perianths, though whorls of threes are not wanting, and seem to point to their ancestral character, which is retained in Romneyee@. Moreover Papaver orientale not unfrequently develops a third smaller sepal, which may or may not be accompanied by a trimerous corolla, which I take to be due to atavism. If this theory be correct, the ancestors of poppies were trimerous ; but by the arrest of one part in the calycine and one in each of the corolline whorls, these have become dimerous and equitant in zestivation. р ety А e rare instances, due to insect agency, such as Polygala chamæbuxus and Phaseolus, which аге, во to say, “ distorted.” І use the word symmetrical here in its popular, not botanical sense, as usually restricted, though Sachs applies it thus. The term zygomorphic has been used to express bilateral symmetry, and, if вепе- rally adopted, would avoid much confusion. THE REV. С HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 181 Or again, if we start with a quincuncial arrangement, the equitant condition is arrived at by suppressing either the first or fifth member. This, I think, has probably been the case with Actea (p. 186) and Adora * (р. 474), as well as with the Crucifere t. Circea (Tab. XXV. fig. 2), however, I am inclined to think owes its equitant sestivation either to an original dimerous symmetry derived from opposite leaves, or else it may be due to “symmetrical reduction" from a tetramerous type which prevails in the same order (Onagracec) to which it belongs. If the equitant sstivation is presumably derived from а quincuncial whorl of five parts, by the arrest of one, the four remaining will probably be wholly equitant at first ; but they may undergo further changes. If it be the result of a convolute estivation, then the remaining pairs will more likely be half-equitant. Thus, if Cireewa be derived from а tetramerous type with convolute petals, as occurs in Epilobium, by the arrest of two opposite parts, then it will be seen how each part of the remaining pair will overlap the other reciprocally, and the resulting zestivation will become half-equitant (Tab. XXV. fig. 2). I would not, however, venture to lay very much stress on this supposed difference in the origin of half-equitant as compared with equitant whorls; for the overlapping of another by the edge of any part may be simply due to some slight advantage in growth quite irrespective of any cause arising from antecedent conditions—as, indeed, may be the ease with the calyx of Papaver, which is half-equitant, though, as I presume, its dimerous character is derived from a tertiary type. 2. Tristichous.—This will be seen to be identical with ап imbricate-proper restivation when the whorl consists of three members only. It is very common in the perianth- leaves of Monocotyledons, occurring in the outer whorl about 40 p. c., and in the inner whorl about 60 p. c., as deduced from the diagrams of Le Maout and Decaisne. This possibly would require some modification ; for I found the convolute æstivation also very common, as in about one third of the flowers of Crocus examined. This kind is repre- sented by those authors as occurring in only four genera. Іп а pound of dates I found the percentage of the inner whorl of the perianth as tristichous to be 82, and as con- volute 18. Тһе outer whorl is gamophyllous, and reveals no succession of parts. The origin of the prevailing ternary arrangement in Monocotyledons I take to be simply due to the fact that it does not arise out of opposite leaves, which latter, as I have shown elsewhere, gives rise chiefly to the pentastichous arrangement (Trans. Linn. бос. sec. ser. vol. i. р. 37), so common іп Dicotyledons; hence, as the usual phyllo- tactical requirements demand three leaves for every projected circle, and the floral requirements demand an alternation of position in the successive whorls, so by these two effects there issues the usual alternation of groups of threes. 2; In Dicotyledons there is a group of orders with а ternary arrangement prevailing amongst them (Magnoliacee, Anonacee, Menispermacee, Berberidacee), associated with a few genera or orders having a large number of imbricated parts in their fioral whorls (Dilleniacee, Calycanthacee, Nympheacee, &c.). When the whorls are in * See Note B, р. 194, on the variations of the symmetry of the flowers of Adowa. t See Note А, p. 191, on the origin of the arrangement ofthe parts of a cruciferous flower. 182 THE БЕУ. С. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. threes in these orders, I believe it to arise as just stated, and аз is more fully explained further on (Note C, р. 194) from phyllotactical considerations; that is, a prolonged spiral, instead of furnishing whorls corresponding to cycles, as in Aconitum, is simply broken up into decussating whorls of threes. 3. Pentastichous or Quincuncial.—This arrangement is one in which the parts of а whorl constitute one cycle of that leaf-order. Each successive member is at an angular distance of 144° from the next in order. It appears to be the commonest in Dicotyledons, obtaining either in one, two, or more whorls of a flower, and, from the diagrams of Le Maout and Decaisne, occurs about 50 p. с. This deduction is quite corroborated by observations on individual plants; for in the flowers e.g. of Primrose % (which shows eight if not more varieties of æstivation) the pentastichous was about 40 p. с. Similarly is it with Viburnum tinus. What I have said upon the origin of alternate leaf-arrangements in my paper referred to above, fully accounts for this plan being the commonest among Dicotyledonous Angiosperms. Indeed I would venture to call it one of the fundamental plans, as giving rise to at least three others, if not more, such as the half-imbricate, imbricate proper, convolute, and possibly in part the equitant. 4. Half-imbricate.—l apply this term to a very common arrangement, which does not appear to have been at all recognized before, but which is nevertheless a very important type of zstivation. It is deduced from the pentastichous by the 2nd member of that kind having one of its edges passing under the 4th, and is not only the first step towards the imbrieate proper, and thence to the convolute, but will also explain the estivation of many irregular flowers, such as the “ papilionaceous," that of the Cesalpinee, and the “ cochlear" t, which, too, can be thus accounted for by a simpler method than is usually applied, and moreover a perfectly natural one, as it exists abundantly amongst floral whorls. j In the percentages deduced from the diagrams of Le Maout and Decaisne, that of this kind does not appear very high (16 p. с.). Ап examination of many flowers leads one to think it should be somewhat higher, as, e. g., іп Laurustinus it amounts to nearly 80 p. с., though in the Primrose it is certainly rare; similarly in Ribes it only amounts to about 10 p. c. With regard to the papilionaceous corolla, if we consider the standard as Хо. 1, one of the keel-petals as No. 2 (of a quincunx), which now passes under the wing, i. e. No. 4, the wing on the opposite side of the median line will be No. 3, while the remaining keel- petal will be No. 5, coherent with No. 2. * The diagram given by Le Maout and Decaisne of the Cowslip (miscalled Primrose), on p. 529, has both the calyx and corolla according to the imbricate-proper sestivation, and not pentastichous, which is much the most characteristic form. See Note D, p. 195. Since this paper was sent to press, my attention has been called to Dr. A. W. Eichler's * Blüthendiagramme, in which he observes that this form is found both in regular and irregular flowers, and he gives a diagram of it for Adnan, р. 270, and Valeriana officinalis, p. 275 ; but һе retains the name “ cochlear." This I purposely avoided, 28 it has only a specific and not generic value. + In the translation of Le Maout and Decaisne the word “ cochleate” is used instead of “ cochlear," which is the correct term : the former is descriptive of a spiral shell, and would be applicable to the coiled legume of Medicago. THE REV. С. HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 183 Le Maout and Decaisne endeavour to explain the structure differently, by conceiving No. 1 to be one of the wings, No. 2 the other wing, No. 3 a keel-petal, No. 4 the standard, and the remaining keel-petal No. 5. "They thus derive it from the quincunx by removing the standard from being entirely within to a position entirely without the rest (fig. 466, p. 87). Similarly with regard to the :estivation of the Cesalpinee as represented by the diagrams of Cassia and Cercis (р. 367), though in these some con- fusion in the explanation is made by those authors; for in describing (on p. 87) these diagrams, they вау,--“ The standard retains its normal position [ż. e. within the others], and the quineunx is properly formed." But in reference to the diagrams given in p. 367, this is clearly not the case. Both in that of Cercis and that of Cassia, if we take the outermost of the two anterior petals as No. 1, then No. 2 (whether the spiral be supposed to turn to the right, as in Cassia, or to the left, as in Cercis) is in neither case external also, as it ought to be for a quincuncial or pentastichous arrangement. If, however, we take, as I propose, the exterior anterior petal as No. 1, then the internal petal as No. 5, both the above kinds fall under this type, which I call half-imbricate. It will be seen that No. 1 is not the same petal in both the papilionaceous corolla and that of these two genera of Cesalpinee. That this difference is of no relative importance will be shown hereafter. It may be remarked here, however, with regard to the standard not being identical with the fifth (internal) petal of Cassia, that the wings of the former lie over the keel, while in the latter genus it is just the reverse. This would seem to show that the papilionaceous resemblance is apparent only and not real. The calyx in Cassia is represented as quincuncial, that of Cercis “ open” and therefore indeterminate. I recommend this interpretation as doing far less violence to the fundamental quincunx than that hitherto suggested; in other words, it is acquired by a simple alteration of position of one edge only of one petal. It may be observed, too, that it is obtained by regarding the parts of the spiral as running in the reverse direction to that according to the former method of regarding it. This type of sestivation will also account for the so-called “ cochlear,” as of the Snap- dragon (Tab. ХХУ. fig. 10, and Те М. & D. fig. 468, р. 87, in which the numbers there given should be exchanged respectively as follows, viz. 1, 5, 4, 3, 2 should be Written instead of 1, 2, 3,4, 5). Thus by retaining the innermost (anterior) petal as No. 5, and by merely shifting No. 2 partially under No. 4, we obtain the sstivation as illustrated by the diagram. Another illustration of the half-imbricate in connexion with the quincuncial will be found in the diagrams of Scerophularineg. The calyx of Scrophularia (р. 585) is strictly quincuncial: the exterior sepal, No.1, is posterior *; but if this be compared first with the diagram of -Linaria (p. 584), it will be seen that the 2nd sepal (the right anterior) now passes under the 4th (or the right), the spiral being right-handed ; and the wstiva- tion of the calyx is therefore half-imbricate. It is similar to the calyx of Paulownia (р. 584); only this spiral is left-handed. But now comparing these with the calyx in the * The diigram is wrongly placed. The left-hand exterior sepal is the posterior one in reality. + On the distinction between right- and left-handed spirals, see below, note, p. 186. 184 THE REV. б. HENSLOW ON THE ORIGIN OF FLORAL Ж8ТІУАТІОХ8. diagram of Snap-dragon, p. 583, the 1st sepal is no longer the posterior one but the left anterior; the spiral is left-handed; and the posterior вера! is the 5th, and completely within the others. The corollas of all three genera have the same estivation, and are constructed on this plan, which I call half-imbricate, having, at least most commonly, the 5th petal as the anterior one. It is rare to find any other part than the 5th in the anterior position of the zygomor- phic whorl, when its estivation is half-imbricate. There is, however, one case figured by Le Maout and Decaisne, viz. the diagram of Streptocarpus (р. 598), in which the 4th petal is the anterior one. It is also thus in Azalea and Rhododendron, in which the 5th petal (always the single petal, which is entirely within the others) is the posterior one. Lastly, in Pelargonium alone all these arrangements may be found. The above shows that no necessity can be seen for No. 1 of a whorl being always the same in position and therefore always homologous. I have already remarked that this arrangement furnishes conditions varying from perfect regularity in the whorls to extremely irregular ones. A few references to the pages of Le Maout and Decaisne will illustrate this. 1. Corollas regular or very slightly irregular :— Pittosporum, р. 248; Rue, р. 317; Parnassia, p. 408; Elder, 481. 2. Corollas partially or decidedly irregular :— Moringa, р. 235; Dictamnus, р. 316; Horsechestnut, р. 357; Cassia, р. 367; Rho- dodendron, p. 516; Henbane, p. 578; Vitex, p. 615. 3. Very irregular corollas :— Pansy, p. 241; Lathyrus, р. 369; Snapdragon, р. 87; Lamium, p. 620. Irregular and zygomorphic flowers are sometimes produced without any departure from the quincunx ; for it should be observed that both the quincuncial and the half- | imbricate sestivations can be divided symmetrically into two similar halves (as repre- sented in Tab. XXV. figs. 4and 5) by median lines. Thus is it with Polygala (Tab. XXV. fig. 11), which somewhat resembles the papilionaceous; but the estivation is simply quincuncial and without any modification. Similarly amongst gamopetalous corollas the diagrams given of Alloplectus, p. 597, as well as that of Achimenes, p. 596, and Adhatoda, p. 606, though possessing irregular bilateral flowers, have yet their petals arranged on the quincuncial plan. Irregular flowers with a quineuncial æstivation would seem to be comparatively rare ; and when it does occur the various parts are not always homologous. Thus in Ramondia, p. 600, Alloplectus, p. 597, and Achimenes, p. 596, it is in the 4th part, while in Stilbe, p- 613, it is in the 3rd part, and in Муоротит, р. 610, it is in the 1st part, which respectively occupy the anterior position. 5. Imbricate proper.—In this sestivation the parts of the whorl may be three in number, as in the tristichous; four, as in Holly, p. 339, and .Euonymus, р. 842; or more commonly five or more. The parts overlap one another in succession, so that the first is wholly exterior, and the ‘last’ (which is really the third of the original quincunx from which it is derived) is entirely within the spiral. Le Maout and Decaisne number the parts in succession, fig. 463, р. 86; but this is not, nor can it possibly be correct, at least in accordance with phyllotactical principles; for if it were, we should have five parts THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL AESTIVATIONS. 185 in asingle projected coil or circle, which сап only arise from the series 1, 4, 2, 1, &c. I observe Prof. Gray also notices the mistake of those authors, which occurs also in other text-books. If, however, we proceed from the last-mentioned ssstivation, or the half-imbricate, in which the 2nd part lies nnder the 4th, and place the 3rd under the 5th, we at once obtain the present kind. In the percentages deduced from the diagrams of Le Maout and Decaisne, it will be seen that this kind is much less common (only 9 p. c.) than the preeeding; and what appears true for genera is also true for flowers of individual plants. Thus Laurustinus and Primrose give the percentage of imbricate proper as only 10. 6. Convolute.—As the last is derived from the half-imbricate, so this is obtained from the imbricate proper, by placing the 1st part under the 3rd of the quineunx ; so that if we write the connexion of these three kinds successively, it will appear as follows :— starting with the quincunx, if we place the edge of No. 2 under No. 4, we obtain the half-imbricate ; next, place the edge of No. 3 under No. 5, and we get the imbricate proper ; and lastly, in addition to these, place the edge of No. 1 under No. 3, and the convolute is secured. As long as the convoluted petals are erect and their median lines vertical there will be no torsion; but in some cases the apex appears as if it were artificially twisted, as in Mallow, Flax, «с. In these cases the word “ contorted” can be applied either with or without “ convolute " as well. In turning to the percentages deduced from the diagrams of Le Maout and Decaisne, it appears that in corollas of different genera the number recorded gives the convolute variety a position of 16 p. c.; and in comparing this with the percentages of the same kind, deduced from flowers of the same species, as of the Primrose, it amounts to 15 p. c., or almost the same. 7. Falvate.—Of this kind of estivation, with its two varieties induplicative and reduplicative, little need be said. They are comparatively rare both among sepals and petals. With regard to the origin of it, I believe, in possibly the majority of instances, it is due to a degradation from the imbricative kinds, using that term in a general sense. It is at least probable in all instances where the flower is presumably due to a spiral phyllotaxis. The diagram of Geum, р. 882, appears to afford an instance of a transition from the quincuncial to the induplicative in the petals, while the calyx has assumed a reduplicative form. I have reproduced this diagram in Tab. ХХУ. fig. 18. In some cases where the flowers are dimerous or tetramerous, a symmetry which has probably risen from a similar opposition in the foliage-leaves *, it is also probable that no spiral arrangement has ever intervened. This may explain the fact of Clematis being nearly the sole genus (Naravelia is properly a subgenus only) of ри which has opposite leaves and tetramerous flowers with a valvate calyx. Similar is it with the genera of О1еасее, аз the Lilac, the sepals of Epilobium, of Rhizophora, &с. Ап illus- СА hope to develop this idea on a future occasion. І would just call attention to the significant fact that of the genera of Бовасеж, comprising 71 in all, as recorded in the * Genera Plantarum’ (Benth. and Hook.), three only have opposite leaves, and these three are alone characterized as having tetramerous flowers, viz. Zhodotypus, Eueryphia, Coleogyne. SECOND SERIES.—BOTANY, VOL. I. Әт 186. THE REV. G. HENSLOW ОХ THE ORIGIN OF FLORAL ZESTIVATIONS. tration of opposite foliage-leaves coalescing in venation in a valvate manner is seen in some species of shrubby Veronice. Of course there are instances of a valvate calyx accompanied by an alternate arrange- ment in the foliage-leaves, such as Malva. These, as also all ** open” kinds of xestiva- tion, may be and probably are due to degradation, in consequence of which the parts only ¢row just sufficiently large to meet in the former case but not overlap, or else, as in the latter instances, to fall short of one another altogether, as in the Umbellifere. V. REVERSAL OF THE SPIRAL. А peculiar feature to be noticed is the frequent reversal of the direction in which the parts of successive whorls run, whether of the corolla in being different from the calyx, or the inner whorl of a monochlamydeous perianth from the outer. Of the 35 diagrams in Le Maout and Decaisne which have both calyx and corolla quincuncial, two have the calyx right-handed and the corolla left-handed. Оп the other hand, four have the calyx left-handed, while the petals form a right-handed spiral. All the rest have the calyx and corolla either right-handed or left-handed together *. In examining several flowers of the same species, as in other matters, so in reversal, I found many single kinds of plants furnishing this phenomenon, as Crocus and Malva sylvestris, in which the right-handed and left-handed contortions were in equal pro- portions numerically. VI. THE NUMERICAL ORDER OF THE PARTS OF WHORLS. In examining the 34 instances of dicotyledonous Angiosperms, where both calyx and corolla are arranged on the quincuncial plan, the first petal of the corolla was found to be in 13 instances between the 1st and 3rd sepals; in 11 instances between the 1st and 4th sepals; and in 1 instance between the 5th and 8rd sepals. In all these cases the imaginary spiral line would be in the same direction, though not continuous from the 5th sepal to the Ist petal. In 9 instances the spiral has to be retraced to discover the 1st petal. Of these latter in 3 instances the 1st petal lies between the 5th and the 2nd sepal, and in 6 instances * With regard to the direction of the spiral, I always signify as right-handed the parts of a whorl the sections of which, as represented by a diagram, overlap on the right side, as seen from the centre, 1. е. corresponding with the movement of the hands of a watch. I find this method more convenient for diagrams than the reverse, which is employed іп the terms sinistrorse and dextrorse for spiral climbers. In that case it is more convenient to regard the spectator as being on the outside ofthe climber. Prof. A. Gray (l. c.) says :—‘ I note with satisfaction that Bentham and Hooker use these terms to signify from left to right or from right to left of a person supposed to stand outside of a closed bud, which is surely the natural position of the observer.” This is without doubt perfectly true for 4 bud, but it is not for a diagram, which is a transverse section of a bud. In this case it is “ natural ” for the eye to rest on the centre. People invariably regard the hands of a watch as going from left to right; yet if instead of Tegarding yourself in the middle of the watch, you remember that уоп are outside of it, then the hands go from right to left when passing from 3 to 9 o'clock. There appears to be some confusion in Le Maout and Decaisne’s work; for they describe their diagrams, р. 80, of phyllotaxis as left-handed, though they are really right-handed if traced from the periphery to the centre; yet the descriptions of the two convolute corollas of Hermannia, р. 284, correspond to the method I һауе adopted; compare also the diagram of Dianthus, p. 256 which, again, is co i ibed i у again, ntrary to the latter, and is left-h though described in the text as ri and is anded, ug. THE REV. б. HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 187 between the 2nd and 4th; but in no instance amongst the diagrams given does it pass beyond two sepals. Lastly, in 7 instances, as stated above, the spiral of the corolla was completely reversed in direction from that of the calyx. Similarly in turning to Monocotyledons we find, of the two whorls of the perianth, that some genera have both revolving the same way, others have one whorl the reverse of the other. Moreover the point where the second whorl begins in either case is as variable as in Dicotyledons. Thus, of those genera where both whorls revolve in the same direction, Pinanga, p. 812, and Commelyna, p. 568, commence the inner whorl with their 156 part between the 2nd and 3rd leaves of the outer whorl, both being right-handed spirals. In Juncus, p. 864, Phormium, p. 844, Lapageria, p. 856, and Alisma, p. 800— all being left-handed spirals—the Ist leaf of the inner whorl lies between the 1st and 2nd of the outer. Again, in Colchicum, p. 850, Fritillaria, p. 843, while both whorls revolve to the right, the inner commences with its 1st leaf between the 3rd and the 1st of the outer. In Butomus, р. 798, Snowflake, p. 786, and Toffieldia, p. 851, the inner whorl com- mences diametrically opposite to the last leaf of the outer (that is, it lies between the 15$ and 2nd leaf of that whorl), and moreover is in these three reversed in its spirai direction. In Asparagus, p. 859, the 1st leaf of the inner whorl lies between the 2nd and 3rd ; while, lastly, in Tradescantia, p. 868, and the Pineapple, p. 766, it lies between the 1st and 3rd. These statements clearly show that there is no single law causing a similarity of order in every case. On the contrary, the two whorls seem most capriciously arranged, the only common character being that in all flowers of two whorls to their perianth these should have their members alternating in position. The preceding having been deduced from the diagrams of genera, Т examined several species, and found that such variations as are here represented in different genera could be all illustrated by the flowers of a single species, as of the Snowdrop, Crocus, &е. I, however, retain the above paragraphs as referring to the diagrams, which will therefore furnish the reader with abundant illustrations. УП. On THE CAUSES WHICH DETERMINE THE VARIOUS KINDS OF /ÉSTIVATION. l. Regular flowers.—If the parts of a whorl are all on the same plane, it appears to be quite uncertain as to which member of the whorl sball be developed as the first or outermost, and that, if a calyx or corolla be both quincuncial, although theoretically their members are spirally arranged, yet practically No. 1 of the corolla may be any one of the five petals, and by no means necessarily as near to 144° as possible from the fifth sepal. This may be due to the fact that as the internodes are suppressed the nourish- ment is not only equally distributed over the circumference of the receptacle, but nearly or quite as equally in time, so that it is, so to say, а mere chance Which part of a whorl gets the first start. Some slight differential impulse seems to occur with the first five kinds of zestivation; but with the convolute, valvate, and open no member сап be called the first at all. i | Аз in a single whorl the order of development of its parts is not always in accordance with the theoretically spiral arrangement, and still less is there any connexion between the : | 2р2 188 THE REV. С. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. whorls by a continuous spiral growth, so also, in the order of development of the whorls themselves, it is by.no means always true that they appear in succession acropetally. Thus, as a consequence of a fiower always terminating an axis, it is the fact that not infrequently there is subsequent intercalation of whorls exterior to some already formed. Thus in Silenec “the first whorl of stamens which appears is opposite the sepals, while the second is produced opposite the petals, and (according to M. Chatin, but my own observations scarcely corroborate the statement) exterior to the first-produced whorls ”%, As far, however, as Stellaria holostea is concerned, the stamens opposite the petals seem to me to be intercalated between the first-formed stamens; so that the andrcecium forms one circle only. The petals, however, are decidedly later than both sets of stamens, the latter having attained considerable dimensions when the former consists of minute semi- circular plates. This subsequent intercalation of whorls is perhaps due to what one may call “developmental energy” being stopped in its axial course, and, thus diverted, evineing itself by lateral or lower outgrowths. "This also helps to explain the acrofugal order of expansion in Adoxa (see Note В, p. 194). 2. Irregular flowers.—There is, at least, a presumable probability, based upon what is known regarding the mutual adaptation of flowers to insects, that all irregular flowers have originally arisen by such adaptations, and that the varieties called “ peloria” are reversions to the regularity of the primitive type Т. Now, in order to become irregular, one obvious process is the greater development of one or more parts of a whorl, often at the expense of the others, which remain more or less dwarfed; so that it would be presumable 4 priori that irregularity would determine which part or parts should arise first from the receptacle. But this does not always appear to be the case, though at present I am not aware that any law has been discovered which regulates the order of growth. The fact that any adult part is larger than another part does not necessarily imply that it started first. Thus is it often with those anthers which appear first, and far outstrip the petals in size, as in Stellaria; yet the petals, when adult, often exceed the stamens immensely, as in the last-named genus and many others. A like remark may be made with reference to stipules of trees, as of the beech, which in the bud grow to four times the size of the leaf to which they belong, yet soon cease to grow, when the leaf ultimately exceeds them many hundred times. М. Chatin (/.¢.) also observes that the polystemonous andreecia of Mesembryanthe- mum, Opuntia, Cactus, &c. are developed centrifugally ; but the inner, older members become partially arrested, and the outer are the first to dehisce their anthers. * “Some Alsineæ,” that author observes, “ with reduced flowers, give rise only to the whorl opposite the sepals. This is the rule in Paronychiacew, where sometimes the abortion of the stamens opposite to the petals is accompanied by the abortion of the petals themselves ? (Bull. de la Soc. Bot. de France, C, 1874, р. 133). А propos of this fact, I would observe that these and other reduced flowers are mostly, if not all, self-fertilizing ; and besides the degrada- tion of the petals, as being no longer required for attracting insects, the number of stamens gets reduced, ог more or less aborted; thus is it in Senebiera didyma, Stellaria media, Viola canina, and other species with cleistogamous flowers ; sometimes the andrccium is even barren, as in some early-developed individuals of Capsella Bursa-pastoris, Glechoma hederacea, &c. t The order of “ emergence” of the whorls of Linaria vulgaris is as follows—calyx, corolla (Ұма is quite —M— —— lastly, the pistil. THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. 189 In Lamium album, as illustrated by ‘Sachs’s Text Book, р. 480, although 3 sepals appear before the other two, yet the adult calyx is scarcely irregular. On the other hand, the 5 petals appear together as equal-sized рарШзе ; the two posterior soon far out- strip the others, then arching over them protect the anthers even from a very young condition. Тһе anterior petal in the adult stage is two-lobed, and might give rise to the impression that the lip was composed of two parts, whereas the history of its deve- lopment reveals the fact that it is really single, while the “hood” is double. This is also corroborated by the fact that two vascular bundles run side by side between the two posterior and shorter stamens. Іп Ajuga reptans a converse process takes place, in that the posterior pair of petals soon become arrested, and the lip now grows rapidly, arches over, and protects the andreecium just as the hood does in Lamium. VIII. VARIETIES OF ÆSTIVATION IN THE SAME SPECIES OF DIFFERENT GENERA RESPECTIVELY. RANUNCULUS BULBOSUS, or Bulbous Ranunculus. Quincuncial to the right. Occurred about 22 p. c. 5 2 „ lef Pa 5 "3975 Half-imbricate » right $ h 6, j % » left 5 5 8 us Imbricate proper ,, right > ЧӘ а is $ » Де »* wc M Convolute ,, » right э $ G а » у Те н » Jg Anomalous :—Corollas having 6, 7, or 8 petals variously folded, but with a general tendency to be convolute; also a few with 5 petals, but abnormally imbricated. VIBURNUM TINUS, or Laurustinus. On a corymb, with 125 flower-buds. Quincuncial to the right. Occurred about 23 p. c. 5 3 „ IBR 55 н Z4 y Half-imbrieate — ,, right Ж DO ETT j ^ „ left 2A XS Imbricate proper , right » н J is о ӨР s in. бт Anomalous :—Corollas with 6 petals, three instances ; with 4 petals, one instance. PRIMULA VULGARIS, or Primrose. Of 120 flower-buds examined. Quincuncial to the right. Occurred about 22 p. c. 33 ” E left РИ 2 27 3 Half-imbricate , right . . . . . 4 examples. is $5 2 mE iuro s Imbricate proper ,, right М i 5 p. c. left m ” 6 ” LE] 29 3 190 THE REV. С. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. Convolute to the right. Occurred about 12 p. с. 33 9 35 left » 33 17 29 Anomalous :— Two instances with corolla 4-merous; two with 2 exterior petals, and 3 eonvolute within; one with 1 exterior and 4 convolute, one with 6 sub-half-imbricate petals. RIBES COCCINEUM, or Crimson Currant. Quincuncial to the right. Occurred about 50 p. c- ээ 33 25 left 35 э? 38 э Half-imbricate — ,, right » Я 6. 33 „ээ 29 left 33 35 6 45 Imbricate proper ,, right й й Go. FUCHSIA COCCINEA. Convolute to the left. Occurred about 67 р. с. Imbricate proper ,, а ӨСКЕН. EIS Anomalous :—1 petal exterior, 1 interior, and 2 opposite and equitant—of this, two examples; one example with six petals, 8 corrugated sepals, 6 stamens, and 4 carpels. Снктвахтнов Снетвт, or Wallflower. The corollas of this plant afford a great variety of wstivations. Тһе chief kinds are as follow :— Imbricate-proper to the right. Occurs about 12 p. c. 33 33 33 left 35 33 14 э? Convolute ,, и right EN лы 33 55 33 left 2) » 6 „э Anomalous :—1 petal exterior with 3 convolute ,, Cae pL ERES Modifications of equitant . . қ i « "OM ri РЕЈ GALANTHUS NIVALIS, or Snowdrop. Of 24 buds there were no less than 14 different ways of arrangement and order together. In the following representations the outer whorl and numerical order are indicated by the upper figures, the inner whorl and the corresponding numerical order of its parts by the lower row of figures. A. Examples of reversal of the spiral in the inner whorl as compared with the outer. В. 1 9 8) Convolute to RSFbSS$ 9-1 i489" t Convolute to L. | me IOS JLE я т m Irc. са B. Examples of both whorls being right-handed :— 1 уа а i 2 4441 82-8 22 d i. 23 8 4 C. Examples of both whorls being left-handed :— o- 94 " 9 3 1 3 2 1 2-19 3e (1 ag 2 convolute. THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. 191 D. Anomalous :—1 leaf external, the other 2 equitant. Crocus. A. Examples of the reversal of spiral in the inner whorl as compared with the outer. R. 1 2 3 В: Д 271.800 oM ing: d 23 2 191Ь,4.- 3:15 Ғата РГ Logos Biv 18. a3 inch Жый 83225, dos itty 49 B. Example of both whorls being right-handed, т 4 3 | "o. goo C. Examples of both whorls being left-handed. 3 Sic Bish odi | 3 2 H арн oM D. Both whorls convolute to the right. E. Outer whorl imbricate proper, inner convolute. Sometimes both to the right, and sometimes both whorls to the left. IX. Note А.--Ох THE SYMMETRY oF A CRUCIFEROUS FLOWER. It has been customary hitherto to refer the structure of a cruciferous flower to a binary or quaternary type, though in different ways * But, seeing that the leaves аге so generally pentastichous, I would suggest a reference to an original quinary arrangement, with a subsequent symmetrical reduction by the arrest of the fifth member of each Whorl; by which it will be seen that the relative positions of the organs as they occur in а cruciferous flower can readily, but of course theoretically, be accounted for. Thus, if the sepals be supposed to be arranged quincuncially, then the external position of the the anterior and posterior sepals will immediately follow, these being numbers 1 and 2 of quincunx. (Compare the figs. 13 and 16 of Tab. XXV.) Next, assuming the Ist of the five petals to lie between the 1st and 4th sepal (for the whorls must alternate), and the lst of an outer whorl of five stamens to lie between the 1st and 4th petals, we shall have the 1st and 2nd retained as stamens (forming the two exterior lateral and lower ones), and the 3rd and 4th (opposite the anterior and posterior sepals) arrested or represented by glands. . The next whorl of stamens arising in a similar manner, it will be seen how the 1st and 4th will form, by approximation, one pair, the posterior, and the 2nd and 3rd the other pair. Lastly, the pistil is probably composed, as has been often suggested, of four carpels; but their union, according to my hypothesis, would be as follows. The two posterior carpels, numbers 1 and 4, coalesce to form the posterior half of the pistil, while the 2nd and 8rd, uniting in front, will form the anterior half. Thus two coherent * A summary of the various theories шау be found in Le Maout and Decaisne’s ‘ Botany,’ pp. 229, 230. See also * Note sur une monstruosité de la fleur du Violier (Cheiranthus Cheiri, І.) par M. P. Duchartre, Ann. des Se. Nat., õme série, xiii. р. 315. 192 THE REV. С. HENSLOW ОМ THE ORIGIN OF FLORAL ASTIVATIONS. margins give rise to the posterior placentas as well as the stigma opposite to them (this latter being composed of two adjacent marginal and coherent stigmas). Similarly the two anterior margins of the 2nd and 3rd carpels will give rise to the anterior placentas, with their combined stigmas. Тһе four lateral margins I conceive as cohering down the middle of the valves; but their placentas and ovules are usually arrested. Тһе stigmas, however, of these lateral margins are normally present in Parolinia. That the stigmas are marginal in the Crucifere was noticed by Robert Brown (Misc. Bot. Works, i. p. 558); so that the anomaly of their position over or opposite the placentas disappears. Fig. 17, Tab. XXV. (copied from Duchartre /.c.), shows these marginal placentas and stigmas developed on a carpellary stamen. Тһе same stamen also shows (fig. 18) how the replum is formed, viz. by a prolongation of the outer side of the margins, the ovules being produced at a short distance within the edge. Lastly, fig. 19 (also from Duchartre) illustrates one of many combinations of carpellary stamens, in which the kind of arrest of the placentas and ovules I have spoken of, and which I conceive takes place along the median line of each valve, is seen at а. It may be ob- served, lastly, that the dehiscence of the siliqua, indicated at * in fig. 16, is, according to the above interpretation, strictly loculicidal, perhaps the commonest of all the methods, and is thus relieved of all the abnormal appearances. There are some fruits of the Crucifere in which the venation is conspicuous, and would seem to corroborate this view. Thus in Lunaria and Iberis, for example, the transverse and branching veins are given off from the margins of the valve, and terminate by anasto- mosing on the median line. In other cases, where the venation is more obscure, a trans- lucent line may be observed down the middle of the valve, marking, as I take it, the line of cohesion. Besides Parolinia, mentioned above, Tetrapoma has four carpels ; and Brassica as well as Cheiranthus not unfrequently develop two in addition to their normal number, under cultivation. With reference to the supposed quinary origin of the flower, it may be observed that Capparis t has sometimes five petals; and J/egacarpea, if the figure given by Le Maout and Decaisne be true, may have exactly 10 stamens. It may be asked on what grounds do I assume the right to suppress the fifth part of each eyele. Тһе reply is, that such symmetrical reduction from a normal and typically higher symmetry is extremely common. Thus Rue (including Haplophyllum), Ааоза, Hypopitys, Potentilla, Tormentilla, &c. have habitually both 5-merous and 4-merous flowers; while the petals of Lythrum Salicaria range from 5 to 7 in number; similarly 4—8-merous flowers are found on Jessamine, Elder, &c. If now we attempt to account for this reduction or increase, it is only conceivable by its being done by some similar or identical arrest or development respectively 1 in each whorl. If, for instance, we draw alternate whorls of fives, and strike out one part, but any of each whorl, it will be seen how impossible it will be to form uniformly alternating whore of fours out of the remaining parts. If, on the other hand, each whorl be t 1 Лааг to о ей as whatever theory he advanced for Crucifere must apply to Capparidec as vell. THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL JESTIVATIONS. 193 numbered quincuncially, No. 1 of each whorl being nearest that of the preceding, and on the same side of it, and then if we strike out all those, say, numbered 5, the re- maining four parts in each whorl, on closing up the gaps, instantly form regularly alternating whorls of fours. Thus, then, I conceive the cruciferous flower to have arisen by а symmetrical reduction from fives to fours—first by the arrest of develop- ment of every fifth part of quincuncially arranged whorls, then by an additional arrest of a posterior and anterior stamen of the outer whorl, as well as a partial arrest ofthe carpels as described above. A botanist who accepts the idea of a Crucifer being referable to a binary symmetry, is M. Victor Meschajeff*. He regards the two stamens constituting one of the taller pairs (as also of two adjacent petals) as reversions to a primitive type, and considers that the usual free condition of the stamens is a result of dédoublement. I would take this opportunity of expressing my belief that this is a fundamentally erroneous view. It seems, too, to agree with Dr. Masters's expression of “ inseparate" as a substitute for “cohesion.” If cohesion, however, be objected to as implying а previous condition of freedom which never existed in petals normally joined, so “ inse- parate" would seem to imply that union was prior to freedom. There could be no reasonable objection, as far as I see, to the term “ inseparate," provided no explanation of the condition is implied, but simply the fact that the parts are “ not separate." Dr. Masters, however, qualifies the term by saying that the parts thus described as inse- parate are not separated. The use of this participle at once introduces а new idea, and implies the existence of some tendency ѓо separate what was before joined. Now this seems quite as objectionable as the term cohesion seems to that author (* Botany for Beginners,’ p. 27) f. But, according to the laws of evolution, integration, where it exists, is always sub- sequent to isolation or freedom. The rare cases of * dialysis," as of Convolvulus and Campanula with free petals, or of ** solution," as of Pyrus and umbelliferous flowers with an inferior calyx, would be cases of reversion ; while, in the Crucifere, freedom of parts, except іп the gynecium, is clearly a fundamental condition, as it has no affinity with any gamopetalous order: consequently the occasional cohesion of petals or stamens in any cruciferous flower is not a reversion but a more advanced differentiated state, and appears to be due simply to the close approximation of the parts thus occasionally joined. ! If such a flower should ever have an apocarpous pistil, thai would be a true instance of reversion ; but no case of union could possibly be such. Vocal For the above reasons I have always looked doubtfully on “ chorisis ” аза principle i flowers. Doubling or an increase of number of parts is common enough, as indeed * Bull. de la Soc. Imp. des Nat. de Moscou, 1872, no. 2, quoted in Bull. de la Soc. Bot. de France, A, nis p. 16. T Dr. Masters has been gcod enough to give me the following comment on these remarks el * Inseparate' I apply to cases where separation of parts, though * potential, has not yet occurred, — — if the course of develop- ment be arrested, it may never actually occur.” But it is this very ‘potentiality, ы tendency to "t parts, Which I can only recognize as identical with atavism or reversion. Hence I think it ought not to be considered a principle of development, but rather а case of retrogression, and therefore exceptional. SECOND SERIES.—BOTANY, VOL. I. 2E 194 THE REV. С. HENSLOW ОМ THE ORIGIN ОЕ FLORAL ASTIVATIONS. occurs frequently in the Crucifere when two stamens arise from the same quadrangular or oval gland instead of a single lateral опе ; but this is not that one has given rise to two, but that in addition to the normal one another is present. If, instead of regarding only the life-history of a plant per se, we regard it from an ancestral point of view, then ** cohesion," as of the gamopetalz, is the most correct term ; for it implies that the parts were originally (that is, in their ancestry) free, but are now, in the existing descendants, coherent; whereas *'inseparate " would imply that cohesion was an antecedent condition, and that there existed an ineffectual tendency to separate the parts of a whorl. That a cruciferous flower has arisen through insect agency I have no doubt, and strongly suspect that the lateral single stamens are specially concerned in intercrossing. On the other hand, there are several of our common cruciferous plants which are self- fertilizing, such as Capsella, Lepidium, and Senebiera. In some of these I have often deteeted the pollen-tubes penetrating the stigma even in unopened buds, as in winter. These I take to be degraded conditions. In these, too, as far as can be judged from the relative heights of the stamens and stigmas, it would seem that the four taller stamens are chiefly, if not solely, concerned in self-pollination. Іп Senebiero didyma it may be observed that there are but two stamens; and they represent the taller pairs, one for each pair respectively. This reduction is a condition not infrequent in inconspicuous self-fertilizing flowers; for the number of stamens is often reduced, as mentioned above, in the flowers of several of the Alsinee which do not require insect agency. X. Note B.—On Adoxa moschatellina. It may be interesting to some to have recorded the varieties which are found in the symmetry of the terminal as compared with the lateral blossoms. Of 71 heads of blossoms examined, all of which had the apical flower composed of whorls of fours, 60 had all four flowers 5-merous below, | 3 had two 4-merous and two 5-merous below, 2 had three 5-merous and one 6-merous below, 3 had three 5-merous and one 4-merous, 1 had four 4-merous, 1 had two 6-merous and two 5-merous, 1 had the apical flower 5-merous, with three 6-merous and one 5-merous below. This plant is also remarkable for the order of expansion in the different parts. The terminal flower opens first ; and its parts expand all at once. Of the lower flowers the two upper sepals of a flower open out first ; then the two upper stamens are mature, and shed their pollen; next, the anthers below them dehisce in succession downwards. during which period the lower sepals begin to expand. DE Note С.—Ох THE ORIGIN OF THE TERNARY AND QUINARY SYMMETRY OF FLOWERS | WITH INDEFINITE AND SPIRALLY ARRANGED MEMBERS. : обрне the two genera comprising this order, both of which have Ор site leaves, Calycanthus illustrates an " гае from opposite leaves to the зг THE REV. С. HENSLOW ON THE ORIGIN OF FLORAL ASSTIVATIONS. 195 arrangement in the bracts or bract-like sepals of the flower, but no differentiation occurs between the bracts, sepals, and petals respectively; whereas Chimonanthus (see fig. p. 191, in Le M. & D.) would seem to be a more highly differentiated type, in that not only is the calyx now distinguishable from the corolla, but five anterior stamens con- stitute a distinct whorl by themselves, and the indefinite barren ones of Calycanthus are here represented by five also. How whorls or cycles of fives arise directly out of opposite parts has been fully described in my previous paper, quoted above, р. 181; so that I need not dwell upon it now. Magnoliacee, Berberidacee, Апопасее, and Menispermacee are orders characterized by a ternary arrangement in their flowers, and by the whorls being usually double. The origin of this might at first sight be thought to be the same as of the ternary arrangement, which is almost universal in petaloid Monocotyledons—namely, to have arisen from the tristichous phyllotaxis. But the fact that the perianth consists of 6 parts (=3+3) in the latter, and twice or more than double that number in some of the former, is significant, and would seem to imply a different origin, which I believe to be as follows :—For any phyllotactical arrangement corresponding to the frac- tions of the usual series, excepting 1 or distichous, every projected circle contains three leaves; and on arresting the internodes of the flowers there always follows the tendency to alternation in position with the parts of the whorls immediately above and below. Hence the “threes” of each circle arrange themselves in a decussating manner. Thus іп Berberis vulgaris we have З bracts, 6 (—3--3) sepals, 6 (=3-3) petals, 6 (23--3) | stamens, making in all 21 parts, corresponding to a cycle of the series 55. On the other hand, a flower of Aconitum, according to Braun, as quoted in Sachs’s Text-Book, р. 530, is composed as follows: the sepals are pentastichous, or $; the corolla octastichous, or 2; the stamens зг, and the carpels 3. Hence it would seem that there can be be at least four methods of disposing the floral members when numerous and spirally arranged, viz. :— 1. Spirally and with no breaks, then mostly з, e.g. Magnolia. ` 2. Broken up into fives, or 2 arrangement . . , Chimonanthus. 3. Broken up into threes, decussating . . . » Berberis. 4. Broken up into whorls of 5, 8, 21, and 8 . . „ Aconitum. XII. Note D.—Owx THE NATURE ОҒ THE COROLLA OF Primula. With regard to the theory of Pfeffer (see Sachs's Text-Book, p. 531), that in the Primulacee the corolla is an outgrowth of the andrecium, I think there is more than one argument against it. Not only would the whole of the Myrsinee have to be interpreted in a similar manner, but all analogy seems opposed to it. | That the stamens are opposite the petals is true; but then Samolus, with its staminodia alternating with the petals, throws light upon this peculiarity; and if we examine the fibro-vascular bundles, it will be seen that ten pass into the tube of the corolla, five larger ones up the middle of each lobe, and five smaller ones are intermediate; each of the larger sends off a small bundle to the stamen adherent to the petal, respectively. Тһе main portion 196 THE REV. G. HENSLOW ON THE ORIGIN OF FLORAL ASTIVATIONS. of the bundle continues its way to the notch at the summit of the lobe, and branches within the lobe. Now this is identically the same as in the calyx of Prunus, in which, as in the corolla of Primula, there are five median bundles, which do not bifurcate until they arrive at the lobes. Тһе five bundles alternating with the petal-lobes of Primula or calycine lobes of Prunus bifurcate also on arriving at the notch between the lobes. Тһеу then send off dichotomously branching members up the adjacent edges of two lobes—that is, of two different members of the whorl. The conclusion I would draw is this—that the calyx-tube of Prunus is homologous with the petioles, the lobes with the blades, and that the same holds good for the corolla of Primula; and I think we are justified in concluding that the corolla of the latter is а veritable foliaceous whorl. Moreover the sestivation of the corolla of Primrose is strictly the same as in many other plants, and, if analogy is to be trusted, confirms this belief. Lastly, what is of chief importance is, that the theory being based, as no doubt is the only sound method, on development, is, however, as I interpret the facts, based on а misunderstanding of them—namely, that because the stamens appear first, and the lobes of the corolla suó- sequently (but they arise as soon as the anther is developed, and are all connected бу а horizontal band), therefore the latter is an outgrowth of the androecium. But, as has been already observed, invariably acropetal or centripetal development of the whorls of flowers does not occur. That the corolla-lobes should develop after the stamens is by no means without parallel; for it occurs in many plants, as in the few following ex- amples, in which the whorls are written in order of development :— Ranunculus acris: Cal., St., Pi., Cor. Stellaria holostea: Cal., St. opp. pet., St. opp. sep., Pi., Cor. Lychnis dioica: same order, but without Pi. Veronica chamedrys: Cal., St., Pi., Cor. Cerastium glomeratum (self-fertilizing): Cal., Pi., St. opp. sep., St. opp. pet., Cor. Arenaria trinervis (do.) : same order. Henslow de: Ы РАРАУЕК half equitaxt que ү, ігчзйойои5 гегогесае prope. 7 o * да -imbricate Hanhart imp ^ U.Bert |+ erjeau hth ^T^ Th тат пс ТАЯ N i3 r L RAL Азіі Wit ГИЯ XII. On some Points in the Histology of certain Species of Corallinacee. By Major- General В. J. Newson, R.E., and Prof. Duncan (М.В. Lond.), F.R.S., Pres. Geol. Soc. Communicated by Dr. MURIE, F.L.S. (Plates XXVL, XXVII.) Read June 15th, 1876. CONTENTS. I. Notice of the Histology of the Corallineæ by Quekett and Decaisne. II. Notice of M. Rosanoff’s Essay on the Histology of the Melobesiæ, so far as it concerns this communication. III. On the Bermuda Corallinaceæ ; their simple and compound filamentous Epiderm; the structure of the external cellular tissue of the Frond and its relation to the deposit of carbo- nate of lime; the nature of the internal tissue. Their physiology. IV. The Histology of the Epiderm of the Frond of Corallina officinalis from temperate seas ; notice of the Structure of the Cells of the Mass of the Frond. V. The Filamentous Processes of the Melóbesiæ of the Bermudas. I. Notice of the Histology of the Corallinee by Quekett and Decaisne. PROF. JOHN QUEKETT, who lectured at the Royal College of Surgeons, London, in 1850-1852, on the skeleton of invertebrate animals, published his course under the title ‘Lectures on Histology,’ in 1854. In the second volume that distinguished micro- scopist, whilst considering the calcareous skeleton of the Coelenterata, very properly distin- guished it from the remarkable hard substance of the Lithophytes, or the plants com- monly known as Corallines. He gave a description of the common Corallina officinalis of the British coasts and also some important notices of the microscopical anatomy of the group. After noticing that when the Corallines are acted upon by dilute hydro- chlorie acid and the coating of lime is removed they become as flexible as any plant, һе proceeded to explain that the terminal branches, which are of a round figure, exhibit the organs of fructification. The other species, Corallina monile and Corallina opuntia, he explained, are both largely coated with lime, but their joints are more evident than those of the common species. He then, after remarking upon the relation of the Nullipores to the Corallines, suggested that the comparison of the skeletons of the plants and of those of the true zoophytic corals is most interesting physiologically ; for in both instances abundance of caleareous material has been separated from the sea-water by a vital process He stated that on making a vertieal section of Corallina officinalis ** we shall find that on examination with the lowest powers it will exhibit two kinds of structures, both of which are essentially cellular—that on the exterior being formed of small cells of hexagonal figure, whilst on the interior they are more elongated and generally of a brownish colour: this is especially the case if the section should include a joint. In the fresh state the contents of the cells can be easily made out, and the central ones are not SECOND SERIES.— BOTANY, VOL. I. 2F 198 MAJOR-GEN. NELSON AND PROF. DUNCAN ON THE unfrequently full of greenish granules like ehlorophyll. Тһе lime is not in the interior of the cells, but appears to be on the outside of the cell-walls, whieh are rendered opaque and thick in consequence." He examined this section with a power of 200 diameters, ` and delineated the results * Тһе drawing shows a very irregular patchwork of white spaces amidst a dark matrix; and the spaces represent the irregular cells without lime surrounded by the calcareous intercellular mass. Some long fibres are shown to be part of the articulation ; and they are destitute of the mineral. Не illustrated the general anatomy of the plant further by making a transverse section of a joint (i e. a part between two articulations), and by dissolving it partly in dilute hydrochlorie acid. Тһе drawings t given by him exhibit the outer part of the section decalcified, and the inner as still retaining some carbonate of lime—so that externally there is a broad series of more or less hexagonal cells in radiating rows, whilst internally the cut extremities of the smaller cells are seen to be surrounded by opaque mineral matter. There isa remarkable addition to this delineation, which is not referred to in the context, in the form of а space included between two lines, which represents a homogeneous tissue on the surface of the lithophyte. In noticing Соға па incrassata, he explained the very flexible nature of the arti- culations—and that by making a vertical section of one the elongated cells of it could be distinguished without any lime upon them, and having often a green colour. Тһе delineation given of these sections shows long internal and hexagonal external cells, with a perfectly plain margin to the free surface of the plant. Finally, in noticing the absence of any such articulation-cells in the Nullipores, he accounted for it by their slow growth in comparison with the more rapid increase of the Corallines, whose calcareous covering was therefore all the less. It will be found on comparing Quekett’s histological work with the essay by M. L. . Decaisne $, that he was so far in advance of that author that it is hardly necessary to allude to it here. But there are some points of interest in M. Decaisne's paper, and some others which, if not explained, might appear to have forestalled the present com- munication. L The illustrations given by M. Decaisne are in plate 17 of the 2nd series of the Annales des Sciences Naturelles, 1842, vol. xvii. (the plate is marked “tom. 16”); and nearly all have been much improved upon by Quekett. But the illustration of Cymopolia barbata is remarkable. The figure (12), of the natural size of the plant, resembles the appearance of the structures of many Corallines when magnified. Fig. 124, however, explains that the endings of the frond аге dicho- tomous branchlets, which are covered with carbonate of lime. They thus differ from what will be explained further on. M. Decaisne noticed (see fig. 14 а) that a mucilage covered the Corallina opuntia (Halymenia opuntia), and that it formed a sort of cuticle which covered the surface of the articulations; but his diagram of the network of cells formed by the contact of those of the superficies is not in advance of that of Quekett. _ * John Quekett, Lectures on Histology, 1854, vol. ii. р. 169 её seq., fig. 85, p..171. T Quekett, op. cit. vol. ii. p. 171, fig. 85, fig. d, and fig. 86. + L. Decaisne, Aun. des Sci. Nat. 2nd ser. vol. xvii. plate 17. Classification of Algw. HISTOLOGY OF CERTAIN SPECIES OF СОВАТЛАМАСЕХ. 199 П. Notice of M. Rosanoff's Essay оп the Histology of the Melobesize, so far as it concerns this communication. In 1866 Rosanoff contributed a paper to the Société des Sciences Naturelles de Cherbourg (2° sér. tome xii. pp. 1-112, plates i.-vii.) on the JMelobesie ; and, apparently ignorant of Quekett's work, he proved that anatomist to be correct, and advanced much beyond him. Аз we do not include the reproduction of those lithophytes in this com- munication, it will only be necessary to notice what M. Возапой states regarding the vegetative growth, so far as it relates to the Corallines. Не considers that the carbonate of lime is in the intercellular spaces, оғ between the longitudinal walls of the cells, in the cell-walls, and in the position of the primordial utricle, and that it is arranged molecularly in the cellulose of the cell-wall. Не does not find it in a crystalline or granular form, but asa thick and more or less perfect layer. He notices the variable amount of the carbonate of lime in different species. Оп examining the cell-walls of Welobesie he found them (before decaleifying) punctate or granular; and this structure resembled parallel lines crossing each other at almost right angles: hence he considers the lime to be intimately associated with the cell-wall structure. Не observes that the repro- duetive parts are deficient in carbonate of lime; but he does not find that the rapidly growing portions of the plants are so deficient in it as might be expected ; nevertheless he asserts that the parts deficient in the mineral are the free surfaces of the terminal cells. M. Rosanoff shows that the cells of the plant are furnished with pores, one on each side of the cell-wall, perpendicular to the direction of growth, and that the pigment, starch, and protoplasm of the cells accumulate in their upper and outer parts. He explains that although each series of cells is radial, and grows by increase of the periphery, there is an increase laterally, and which relates to the shape of the frond. This takes place by what Ве calls copulation of cells, and is influenced by the pores in contiguous cell-walls. Тһе copulation appears to be an absorption of the cell-walls of contiguous cells and a resultant irregular and large cell; and the pores permit of the nutrition of the deeper cells, so as to contribute an amount of common nourishment. He describes some large cells termed heterocysts. M. Rosanoff explains that the Melobesie have a glutinous feel when touched on their surface; and in his diagrammatie sketches he places isolated cells on the free surface, each being the termination of a series leading deeply down *. The Corallinaceze have been the subject of our independent investigation ; and most of the work of one of us (Major-General R. J. Nelson, R.E., then Lieut. Nelson, R.E.) was done in the Bermudas many years since. The drawings from life were then taken. Subsequently one of us working on the subject of the physiology of the Litho- phytes, and unknown to the other, had examined the British and some other Corallines With care. Hence this joint memoir. _* Other writers (who do not offer any thing new) on the histology of the vegetative struetures are Harvey (Brit. Marine Alg. pl. 13 С; Phys, Brit. pl. 222), Decaisne (Ann. des Sci. Nat. 2° вёг. xvii. pl. 17. fig. 1, xviii. P- 119; The Micrographic Dictionary, 3rd edit.). See also Lamouroux, Bull. Soc, Phil. 1812 and 1816, p. 313; Philippi, in Wiegm. Arch. 1837; Kützing, Phyc. Gen. and Spec. Alg. i 2 F2 200 MAJOR-GEN. NELSON AND PROF. DUNCAN ОМ THE III. On the Bermuda Corallinacee ; their simple and compound filamentous epiderm ; the structure of the external cellular tissue of the frond, its relation to the deposit of carbonate of lime, and the nature of the internal tissue. Тһе high temperature of the sea at Bermuda, the absence of a cold season, and the abundance of earbonate of lime in and about its reefs, favour the growth of the Coral- lines and Nullipores. Under these physical conditions these plants assume an appear- ance of great beauty ; for in the still pools amongst the reefs they may be seen in luxuriant profusion, each frond being enveloped in a delicate halo of filiform vegetation. Instead of the white or purplish smooth exterior which is so familiar to the investigator of northern forms and of dead specimens, that of the living Bermuda Corallines is hirsute, the hard calcareous analogue of the less-favoured species of Europe being beset with vigorous cell-growth. (Plate XXVI. figs. 2, 11, 14, 15; Plate XXVII. figs. 11, 13.) In the warm sea of the Bermudas white Corallines are rare, and the prevailing colour is green, as in Corallina tuna, or full bright grass-green, as іп Corallina tridens; some- times the tint is bluish-green; and some Corallines are brownish green or even plum-coloured. Their filamentous vegetation, or cell-growth, varies in amount and in its details accord- ing to the species ; but in every instance it is continuous with the outer cells of the mass of the frond. The filaments are wonderfully varied in form * and structure: sometimes they are merely tubular and unicellular; sometimes they are very distinctly articulate ; and in some species they branch. Their length varies also; for whilst in some vigorous forms they exceed in length the diameter of the joint of the frond to which they are attached, in others they are not quite as long, and in certain species they are not more than y5 or зо of the length of the diameter of the plant. Many contain extremely minute spherules of chlorophyll or, rather, of colouring matter, whose minute bodies have such a significant appearance that they suggest some other physiological bearing than that of simple colouring matter., (Plate XXVI. figs, 3, 12, 16.) These filaments, radiating on all sides and crowded together, resembling a confervoid growth on the frond, do not appear to be contractile ; for they are passive under the | movements of the numerous minute animals which move up and down and in and out amongst them. They possibly form the food of many minute animals, and are thus liable to be removed from their parent frond, a process which washing and drying | will imitate. Nevertheless, if the algz are carefully collected and placed on paper, traces of the filaments will last for a long time. Very long filaments which are single, and һауе a direction rather oblique to the frond, are seen in Corallina vermicula. Filaments which are crowded and short, and which bifurcate, exist in Corallina tridens; and still shorter ones are observed with bifurcate ànd club-shaped ends in Corallina tuna. | ‘Under а magnifying power of 100 diameters the outside of the dried frond of Coral- y In order to see them to perfection the plant must never leave the water, and should be put into a glass vessel under the surface, and then removed in its temporary aquarium. HISTOLOGY OF CERTAIN SPECIES OF CORALLINACE. 201 lina vermicula is covered with very regular and small polygonal spaces. Each space is separated by a narrow raised border from its neighbours; and the surface within the border is more or less depressed and has a central opening. Тһе border and depressed surfaces aré of carbonate of lime; and the first represents the junction of the outer cells of _ the cellular tissue of the frond at the outside, and each of the depressed surfaces consists of the carbonate of lime in the cell-wall of a collapsed superficial cell. Where there was no carbonate of lime in the central spot there was perfect cell-wall; and this disappears during desiccation. But the filament, long and slightly oblique, arose from this central spot, and therefore was much narrower than the whole cell from which it sprang, and was continuous. (Plate XXVI. fig. 17.) In some parts of a specimen the minute dot-like space uncovered by carbonate of lime is not central, but close to the edge of the polygonal area; so that the filament did not arise invariably centrally, but often at one end; and it appears to be oblique. The polygonal spaces moreover differ in size, and some are more elongate than others; so that the distribution of the filamentous growth is not perfectly uniform. ‘Towards the end of the frond the calcareous element does not predominate as it does elsewhere; and the green tint of the cells which are not affected by the mineral is very distinct. On decalcifying specimens it is observed that the calcareous matter is granular, and that it is between the cells, in the cell-wall, and also within the cell to a certain extent- The inside of the frond is eminently cellular, and consists of series of long tubes and of compressed globes of carbonate of lime with faint central hollows placed longitudinally andradially. Beyond them is the outer coating of polygonal cells, which are intruded on by carbonate of lime—on their sides perfectly, but on their distal and proximal ends imperfectly, so аз to leave some cell-wall in those situations. (Plate XX VI. figs. 15-17.) Corallina tridens, when dry, has its surface covered with minute pits separated by ridges so as to assume a polygonal shape; and at the bottom of them, even in very old specimens, а dry slender filament may be seen projecting from a central part where there is no carbonate of lime. The spaces between the projecting ridges (or the pits) are pro- duced by the collapse of the cell-contents during drying, and form part of the free ends of the external layer of cells, Each cell corresponding to a pit was infiltrated with car- bonate of lime (except а minute central spot) during life. A very delicate cell-membrane covered the pits and ridges, and was continuous with the filaments also; this will be noticed further on as the analogue of the superficial epidermic layer of the British Corallines. Relics of this membrane are seen in dried specimens. It contained green colouring matter ; and where it has been abraded the carbonate of lime of the underlying cells is glistening and white. The filiform processes are very numerous in the living specimens, and forma complete down on the surface. They arise by a kind of neck-like constriction from the most external cells (those of the pits) of the frond, and are wider at their distal end, being otherwise cylindrical. They give origin at their remote end to two neck-like processes, which subdivide rapidly, and end in long club-shaped cells ; so that the original filament produces four others. "They are all tense with fluid and with marked distinct and nume- Tous globular granules, more or less tinted green. (Plate XXVI. figs. 10-13.) 202 ` MAJOR-GEN. NELSON AND PROF. DUNCAN ON THE Corallina tuna, a bright green succulent Coralline, resembling Corallina opuntia in its habit, bas its broad joints covered with a down of crowded and articulated fila- ments. Тһе filaments are of several kinds. Some are simple long club-shaped cells; others are of the same shape at their base, but become compound by giving forth cells | from the free end ; and many, after having had these secondary cells developed, have others growing from them. All the cells are narrow at their origin, cylindrical, and largest at the end. (Plate XXVI. figs. 1-3.) The surface of the joints of the frond is covered with a very minute areolation. Each minute area has its limiting ridges, and is regularly hexagonal, the ridge between the areas having a fine line on it. Each area is the top of a cell; and the ridge represents the intercellular space and the cell-walls also: it is caleareous. The calcareous deposit does not transgress over the front of the outside cells (those forming the areas) as it does in the Corallines already mentioned; so that almost all the area is semitransparent in dried specimens. But there is a little calcification of the outer wall nevertheless, except near the centre whence the filiform appendage of the cell arises; for the wall has a vitreous appearance. (Plate XXVI. figs. 4-6.) Тһе green colour of some of the dried joints is partly due to the preservation of an external superficial cell-membrane which covered all the calcareous cells, and which, as in the other Corallines, is a kind of lateral expansion of the base of the filiform pro- cesses, and is seen in Corallina tridens (see page 201). Іп the dried specimens the pro- cesses may be just traced, their proximal ends lying in the hollow of the area. Тһе hollowing of the area depends upon collapse of the cell during desiccation. "This Coralline is only caleareous externally ; for within there are ramose masses of great elongated simple cells, forming trunks near the articulation, and very fine branchiets near the underside of the outer expansion of the joints of the frond. Hence, on removing the film of caleareous cell-matter which forms the outside of the joint, these ramifications become visible. (Plate XXVI. figs. 1, 2, 9.) à; On decaleifying a dry specimen, the paucity of the carbonate of lime in the frond becomes evident; for the ridges between the polygonal external cells are then seen to be mainly composed of true cell-wall: hence the mineral is mainly within that structure, and not much between the cells. There has been a deposition within the cells; for the relics of the cytioplasm are seen in some crowded up in a mass. The cells themselves look crumpled more or less, so as to give the appearance of raphides within them. On the edge of a frond, in a decalcified specimen, the bases of the filiform processes could be seen, and a perfect process also. The cell-wall of the processes is well developed, and has glandular markings and slight prominences on it. (Plate XXVI. fig. 8.) The ramose cellular structure within the joints is very remarkable ; some of it consists of huge long cells constricted here and there, and of smaller cells with neck-shaped processes terminating in round bottle-shaped ends, which underlie the external layer of polygonal cells. (Plate XXVI. fig. 9.) ; | These species may be taken as the types which explain the external construction of the Corallines, and, to a certain extent, their internal configuration. They are all rapid HISTOLOGY OF CERTAIN SPECIES ОҒ CORALLINACEJE. 203 growers in comparison with the species of temperate seas; and they contain less carbo- nate of lime, joint for joint, than the well-known Corallines of our seas, although the whole plant may contain much more. Their cells are larger, and the whole plants present in their minute anatomy a great resemblance to some other little-known Bermudan Algee, but which, whilst they resemble filamentous Rhodosperms, include carbonate of lime within their cellular tissue. The function of the filamentous processes and of the external delicate membrane can be readily interpreted. They are the only means by which absorption of the sea-water, and therefore of nutriment and carbonate of lime can take place. When once the intercellular spaces and the cell-walls have become infiltrated with granular carbonate of line, no further nutrition can proceed. At this time the deposit is free to collect within the cells; for the cytioplasm contains it. When large cells are cut into, they emit a rice-water-looking liquid, which deposits granular car- bonate of lime. The fronds increase, as it were, exogenously, by division of the external cells of the layer which gives origin to the filamentous processes. The superficial epidermal layer also develops a succession of external coatings to the frond; and therefore several of the above-mentioned layers are derived from it. The arrest of growth and the deposi- tion of carbonate of lime have evidently some common relation. Hence in all probability the introduction of the mineral in the soluble form of the bicarbonate into the cytioplasm, cell-wall, and intercellular space progresses during the early life of the frond and whilst growth is proceeding. But as starch, cellulose, and carbonaceous granules accumulate, as growth becomes less vigorous a part of the carbonic acid of the bicarbonate is decom- posed, part is retained, and part is lost, and the insoluble carbonate of lime is precipi- tated in the granular form. The merging of one cell into another (the “copulation” of Rosanoff) has not been observed in the Bermuda Corallines ; nor has any thing like a regular porosity of the cell- walls. But the presence of round spots where the cell-membrane is thin and more per- vious to light is occasionally to be noticed in decalcified specimens. In the cell-wall of the long internal cells there are markings which indicate a differ- ence in the thickness of the wall; and on the surface of the frond the delicate external epidermal layer may be distinguished by curious puckerings and markings in its delicate tissue. IV. The Histology of the Epiderm of the Frond of Corallina officinalis from temperate seas; notice of the Structure of the Celis of the Mass of the Frond. The common British Corallina officinalis should be studied in the early spring time ; for then its origin, from a greater or less expansion, may be examined, and there is then not much carbonate of lime in the frond. The expansion and short stumpy fronds are of а beautiful.claret or port-wine-and-water colour, the articulations and the tops of the joints being greenish or colourless. Care should be taken to remove the Stone to which the plant is attached without any washing movement ; or the delicate external layer will be destroyed. The microscopic specimens, sections, &e. had better be mounted in glycerine; and if any acid be used, it should be weak hydrochloric 204. MAJOR-GEN. NELSON AND PROF. DUNCAN ОМ THE acid, the glycerine being used after washing as the medium. A polarizing apparatus is necessary. (Plate XXVII. figs. 1—9.) Let the acid solution act slowly and for some time if the internal structures are to be examined; but if the delicate film of tissue which covers the frond is to be shown, а rapid development of gas-bubbles is rather an advantage; for it is tolerably certain to happen that some bubbles escape and blow out this tissue so that it can be examined. Young specimens should be obtained; for experience and observation prove that minute Gasteropoda browse over the Corallines, and destroy much of the superficial cell- growth before it is much infiltrated with the carbonate of lime. The surface of the frond of the Coralline, and also of the little basal expansion from which it rises, is covered with an exceedingly delicate cell-membrane (Plate XXVII. fig. 3). This is usually demonstrable after the action of dilute hydrochloric acid upon the frond; for the bubbles of gas escaping, separate it from the cellular mass of the plant which it covers. It resembles a beautiful pavement epithelium; its broad and shallow polygonal cells usually, but not always, touch by thin walls, and the whole is colourless. When seen in section in semi-decalcified specimens, the cells are observed to overlap those below, and, whilst covering the intercellular space, to be in organic connexion with the distal ends of the cells of the mass. No calcareous element enters into the composition of this superficial structure. Here and there an additional cell is seen on the free surface and covering one of those polygonal in shape; and near the joints, in some rare instances, the polygonal cell gives rise to a thread-like cell of about the length of 1 of the breadth of the joint (Plate XXVII. fig. 2). In some forms from Port Natal, which greatly resemble Corallina monile, Elis, the surface-membrane extends from joint to joint, covering the articulation and its elongate cells. "This ex- ternal tissue gives no evidence of containing carbonate of lime; but the cells have minute refractive dots in them. Finally, it is thickest at the free end of the joints of the frond, and there often exhibits rather tall cells, united together by a structureless protoplasm. | Beneath this superficial tissue are the terminal cells of the mass of the frond ; and they are continuous with long rows more deeply seated. Each of these terminal cells is ofa light green, and it succeeds a series which, gradually becoming darker with depth, assume a port-wine colour. "The cell-wall is stout, a portion of it being perfectly transparent even in old specimens. "This transparent portion forms either part or the whole of the distal end of the cell ; and nearly all the rest is opaque. On decalcifying, it is found that the cell-wall, especially at the sides of the cell, and not at the ends, is incomplete here and there, as was noticed in the Velobesie by Rosanoff. There appears to be a pos- sibility of connexion between the sides of contiguous cells. That this is a normal con- dition is probable from the small number of the pores; for did the incomplete condition depend on the removal of raphidic masses of carbonate of lime, there would be more of the pore-like markings, and their position would be irregular *. * A careful examination proves that some markings on the longitudinal sides of the cell-walls of Corallines, already noticed, from Bermuda, are spots of excessively thin cell-wall. They are circular in outline, very few in nabor, and transmit light more readily than the rest of the tissue ; they are, then, not absolute pores, but are p like, and are the seat of a very thin and diaphanous membrane. HISTOLOGY OF CERTAIN SPECIES OF CORALLINACEJE. 205 The terminal cells contain viscid protoplasm, dark granules, and a nucleus. Their sides are, after decalcifying, evidently in contact; and when the process has been only partially carried out, а granular mass of carbonate of lime is seen to hide the side walls of contiguous cells, and to intrude on their internal space. Having their side walls in contact after decalcifying, it is not probable that there ean have been much of the calcareous element between the cells; there was some in the cell-walls, and also in the position of the primordial utriele. There is, however, frequently a small amount of caleareous deposition in the distal end of these outer green cells, but not enough to cover the whole surface. Nevertheless, when the whole frond of Corallina officinalis has matured, the whole of the distal end is covered, and. thus the organie connexion between the superficial tissue and that beneath is destroyed. The proximal end of the outer green cells is in close contact with the upper wall of the cells beneath; and in the early days of the growth of the frond there is very little mineral deposition there: but later it occurs, so that finally these cells, and the others beneath them in the mass of the frond, are quite surrounded by opaque carbonate of lime, Which is, moreover, amongst their cell-walls and within to a certain extent. The superficial epithelial-looking layer of thé British Corallines, and of some from Port Natal, is the homologue of the delicate homogeneous layer covering the more rapidly growing and less calcareous Corallines of the Bermudas; and the flat cells of the one are the analogues of the long slender filamentous processes of the others. In some decalcified specimens from Bermuda the cell-wall of the cells supporting the processes, and of the deeper cells also, is very thick, and no pores are to be seen. After much, but not all, of the carbonate of lime has been removed, it is seen to be granular, to be ineluded in the substance of the walls, between them, and in the cell. And when this last state occurs all the solid contents are crowded up in the middle. In a young frond of Соғай па officinalis these outer cells are seen to be dotted on their sides with minute dark punctations after the action of acid. Beneath the outermost green-coloured layer of cells of Corallina officinalis are others of the same colour or else of a light port-wine tint. They have dark granules, some of considerable size, within them; and in the instance of the port-wine-coloured cells the colouring matter is in solution as well as in the granules. The cells are cubical or oval in outline, some being elongate; they are in close contact with those in front and behind ; but an intercellular space exists around the sides. That is to say, after decalcification, the slightest pressure will separate the sides of groups of cells; and it is in this position that the bulk of the calcareous element is deposited. This imperfect adhesion of ше sides and the decided organic connexion between the ends of the cells in linear series, 18 more or less seen throughout the frond. In young fronds the cells, especially when ү E v aggregates ii enisi masses so as to exhibit a small moniliform series; and in the midst is a rather large nucleus and nucleolus. Sometimes кене is to be observed a large circular dark line encircling a small zone of diaphanous wall, in the midst of which 48 а circular dark spot. All these cells are distinctly tinted. (Plate XXVII. fig. 7.) | SECOND SERIES.—BOTANY, VOL. I. ‘they are elongate, are crammed with 26 206 MAJOR-GEN. NELSON AND PROF. DUNCAN ОМ THE Deeper still in the frond the linear series of cells, which are placed radially, are seen. to branch; or, rather, two or three series unite deeply in one, and the size of the inner cells increases. АП are very distinctly separate by their sides after the action of acid. Amongst these deeply seated cells there are some which are long and fusiform ; they run longitudinally through the joint, and deeply. "They are observable in the middle joints, and not in the younger and terminal ones of the young frond. Each is crammed with two or more linear series of slightly flattened spheroids of colourless matter, very amyloid in appearance. And whilst in some a mass of dark port-wine-coloured cytioplasm may exist at one end, and a series of the amyloid bodies at the other, the majority of these cells, which sometimes bifurcate, are colourless; and in many no cell-wall can be distin- guished, on account of the close packing of the contents. (Plate X XVII. fig. 9.) These long cells, with their remarkable contents, are in relation with the filiform cellular processes of the articulations, and are often distinctly continuous with them, especially with the deeply seated ones, which run in a longitudinal direction throughout. Each articulation consists primarily of a mass of long cells placed between the joints, and it has a deep and a superficial part. The superficial part of the articulation consists of a thin layer of epithelium-looking cells, well seen by polarized light, and which is continuous with the common superficial layer of the frond; then follow long cells crowded closely together, and having a faint green tinge. Each cell has a thick wall, which is more transparent than the elongate protoplasmic contents; and usually the cell is long enough to extend from one joint to the other. It then is fixed on to a broader and shorter cell, whose direction is oblique and more or less radial; and it contains more or less colouring-matter, green or dark red. This obliquity of the outer layer of the articular cells close to the joint produces decom- position of the polarizing ray, and the phenomena of alternating light and darkness and of complementary colours. The deeper cells of the articulation are elongate, like the others; they are very long, numerous and crowded, and have thick cell-walls and a linear mass of cytioplasm. They pass straight on in either joint, and merge into long cells without colouring-matter, or else into a series of the fusiform cells noticed above. ! It appears, from examining very young specimens, that the outer part of the articula- tion is developed from the outer layers of the joints by elongation of their cells, and that the deeper layers are produced from the elongate and longitudiual series of the frond. As growth proceeds, the fine tissue on the articulation is lost, the long cells increase in breadth and length and in the thickness of their walls, their contents becoming semi-solid, Occasional dissepiments occur in the long cells of the articulation, and also short pro- cesses of the cell-wall; but often they are not visible. Polarized light, however, is à great assistant in the investigation. A small amount of carbonate of lime is found amongst the superficial layer of cells; but it does not appear to exist in and amongst the deep and long cells of the articulation- Іһ older fronds, and in some Corallines from Port Natal, the superficial epidermic layer is very distinct; and beneath it the cells of the first layer of the mass of the frond HISTOLOGY OF CERTAIN SPECIES ОҒ CORALLINACE Ж. 207 are convex outwards, having very thick walls; but their walls are not usually perfect in outline. They have a double contour-line; and this ceases over а greater or less part of the position where cell-wall might be expected to be found. This absence is due prin- cipally to collapse of the cells after decalcification, and, to a certain extent, to the presence ofa lateral pore. (Plate XXVII. fig. 5.) The absorption of the adjoining cell-walls of cells and the conjugation noticed by Rosanoff have not been observed in the sense in which the terms are employed by that author. In the mass of the frond a large cell may be seen to have incomplete dissepi- ments at its distal end; but these are the beginnings of the sprouting of two perfect cells from it in order to multiply the series. This occurs largely all over the joints, and even in the outer layer, the increase being lateral there and in the direction of the periphery. V. The filamentous Processes of the Melobesi:x of the Bermudas. Like the Corallines, the Jelobesie live and flourish luxuriantly in the Bermudas; and it is interesting to announce that, like their artieulated allies, they present ex- ternal appearances of which no conception has been had by the observers in temperate climates*, The Nullipores chosen as the types upon which to explain the appearance of luxuriant external filamentation, and which form the subject for our plates, are two in number :— First, a very ramose species, with rather expanded ends to short and widely diverging branchlets. (Plate XXVII. figs. 10, 11.) The branchlets are solid, very concentric in their layers, and are minutely punctate. The fine hair-like growth from their external surface is wonderfully developed ; and the extremely delicate and greatly crowded filaments are as long as the branch is broad. They radiate in all directions. The second form arises from a stem ; and its branches, irregular in shape, inosculate, and ascend close together, so as to form a conical mass. (Plate XXVII. figs. 12, 13.) The crowded filamentous growth is placed around the branches, and extends outwardly from the frond to about one half of the diameter of the branch. It is evident that these filamentous growths are the analogues of those of the Coral- lines, and that all have the same important function in the nutrition of the plants. * Nullipores are closely allied to the Corallines both in structure and in their office in nature. Any one жын studied the blocks and masses of coral reef which the Sapper and Miner has blown up when they impede navigation, —or on which the naturalist has crawled, or round which he has been wading or swimming by the hour together— vilae dily admit the good service done, not only by Nullipora, but by many and many с йш ced 25525 Vegetable Coral as а cement, cooperatively with the bond, at times, supplied by Gorgonide „ ш $ E im on : 80-called Coral Reefs, іп а way quite distinct from the comparatively homogeneous proceedings in the structure Tubicola геев. —R. J. N. 269 208 MAJOR-GEN. NELSON AND PROF. DUNCAN ON THE DESCRIPTION OF THE PLATES. PLATE XXVI. . Corallina tuna: natural size. . À joint magnified, and showing the hirsute surface and the long branching internal cells. . Filiform processes or hairs from the outside of Corallina tuna, magnified. . А portion of the surface of an old and dry specimen of Corallina tuna, magnified, showing the spots in the cells where there is no carbonate of lime. Fig. 5. An external cell of the same Coralline, more highly magnified. The dark spot is the position of the former attachment of a hair or filament. Fig. 6. An external cell collapsed and the relics of a dried filament, magnified. Fig. 7. External cells of the frond of Corallina tuna decalcified, and magnified. Fig. 8. Part of the edge of a decalcified frond of Corallina tuna, showing the hairs and glandular markings on them, magnified. Fig. 9. The large simple and branching cells of the interior of the frond which terminate at the bald of external cells, magnified. Fig. 10. A portion of the dried fronds of Corallina tridens. Fig. 11. A portion of a fresh frond, magnified, showing the superficial hirsute covering of filamentous growth. Fig. 12. A group of hairs arising from the external cells of the frond of the same Coralline, magnified, from a fresh specimen. Fig. 13. The appearance of the external layer of cells, with vestiges of dried filaments and of parts with- out carbonate of lime, magnified, from a dry specimen. Fig. 14. A group of Corallina vermicula, after Nelson and from the living specimen. Fig. 15. A fresh frond, magnified, from the life, showing the oblique and long hairs, of Corallina vermicula. Fig. 16. A group of external cells and their long hairs, from a fresh specimen, magnified. Fig. 17. External cells of Corallina vermicula with broken cell-wall, indicating rupture of the filaments. From a dry specimen, magnified. rj е-е N 03 03 С > бо t2 к Figs. 2, 3, 11, 12, 14, 15, 16, аге Жо drawings from nature by Major-General Nelson, R.E.; figs. 1 and 10 are from specimens collected by Major-General Nelson. Figs. 4, 5, 6, 7, 8, 9, 13, 17, are drawn from specimens prepared by Prof. Duncan, and under his direction. Prate XXVII. Fig. 1. A portion of a young frond of Corallina officinalis, magnified. Fig. 2. A portion of the same frond close to the edge of an articulation, showing scattered and long з 5555,:,:8mple hairs, magnified. Ў Hon > А portion of the surface of a | frond decalcified. The lower arched portion has become separated from the upper and darker part by accident ; it indicates the superficial layer of cells. Magnified. ` ВЕ 4. Тһе superficial cell-layer, from above, magnified. ~ 5. The The appearance presented by the outer cells of the frond beneath the layer above (figs. 3 and 4) _ ш decalcified құсалы. The thickness of the cell-wall and its incomplete condition are shown. | Nelson& Duncan, del TRANS.LINN.Soc. SER.2 Bor VoL 1 TAB 26 | 1 | | : Hanhart imp. CORALLINES & STRUCTURAL DETAILS. Nelson& Duncan, del . TRANS.LINN Soc SER 2 Bor VoL 1 Тав 27 CBerjeau hth. Hanhart imp CORALLINES anp NULLIPORES. HISTOLOGY OF CERTAIN SPECIES ОЕ СОКАНЫМАСЕХ. 209 . 6. Internal cells of а frond decalcified, showing pore-like markings, magnified. . 7. Internal cells, crowded with dark-red granules and spheres of the same tint, ending in elongate cells, forming part of the articulation, magnified. . 8. One of the broad cells, magnified, showing a collection of cytioplasm. . 9. A mass of refractive granules from one of the cells in a young frond, magnified. ‚ 10. A portion of a ramose nullipore (Melobesia) from Bermuda, natural size. . 11. А Ше branch of the mass, magnified, showing the great development of the external filamen- tous growth. . 12. An undescribed species of Melobesia. . 13. A portion of the mass covered with the filamentous growth from the external cells, magnified. Figs. 1—9 are drawn from specimens prepared by Prof. Duncan. Figs. 10-13, are from drawings by Major-General Nelson. ХПІ. On the Minute Structure апд Mode of Growth of Ballia callitricha, Ag. (sensu latiori) By W. АксКЕв, F.R.S., M.R.I.4. Communicated by У. Т. TutsEUTON Dyer, MA., FLS. (Plates XXVIII. & XXIX.) Read June 15th, 1876. SOME small fragments only of this alga (mixed with some of the freshwater algæ enumerated by me in the Society's * Journal, No. 87) were at first met with in gatherings made by Mr. Moseley, of H.M.S. ‘Challenger,’ in Kerguelen's Land. They were much macerated and wholly deprived of colour, very few of the cells indicating the presence of contents, and the cellular structure being rendered hyaline. I submitted such a specimen to my friend Prof. E. Perceval Wright, who, aided by one at same time given to me by my friend Mr. Gilbert Sanders, identified itas a form of Ballia callitricha ; and on comparing it with perfect examples in the herbarium of Trinity College, Dublin, the diagnosis was sustained. However, on a closer search of Mr. Moseley's collection, I afterwards found that it contained, in addition to the isolated and broken-up fragments, some complete and characteristic examples of our Ballia; so that the identification was complete. I was much pleased to find these latter in the gathering, as they afforded, of course, much more suitable subjects for study—though even the first small shreds which I had noticed not only presented examples of a minute and apparently characteristic structure within the cells, which thus seems to possess some histological interest, but they likewise showed some interesting points relating to the outward configuration and mode of growth of this alga, all of which, so faras I can see, seem to have hitherto remained unnoticed. The fact that a Rhodosperm should have occurred in a collection made from fresh water and containing characteristic freshwater forms, possibly does not in reality possess that amount of interest which at first glance might seem to appertain toit. It is true that the aleze of marine type which strictly belong to the fresh waters are very few, and they are but of small dimensions and humble habit, and not at all striking. Such аге, for instance, Chantransia, Bangia, Hildenbrandtia. Та Ballia we have a type, indeed, more elevated, more pronounced, than any of these, an elegantly configurated and compoundly branched form ; but I should be much inclined to suppose our specimen was à mere marine waif (possibly blown in by the wind); or it may have been taken from а site to which the sea had only occasional access (and cast in by the surf)*. The more “Таш further indebted to Prof. Perceval Wright for drawing my attention, however, to a record by Dr. Мов- tagne (addressed to the Academy of Sciences, May 13, 1850) of the fact that М. Lemprieure had eolleeted in the running waters of Guiana three species of Bostrychia, one Gymnogongrus, and two of Ballia. These were found either in rivulets of the mountains of Mahuri (** 80 kilo." from Cayenne) or in the watercourse of the ereek Gravier, in the Kan mountains (“40 kilo." from the sea, and at an elevation of 100—150 metres) The waters presented no : trace of salt; and their elevation precluded all thought of any flow back of the tide bearing germs of these plants. Of _ the Ballia examples, which were minute forms, one (Ballia pygm«a, Mont.) grew attached to а Batrachospermum, the other (B. lemprieurii, Kütz.) to stones. 212 МВ. W. ARCHER ON THE MINUTE STRUCTURE AND important interest of the specimen, then, attaches rather to its being the first in which the specialities alluded to have been noticed. The whole plant is of a rigid and tough character; nor, when it is viewed under the microscope, is this surprising, when the cell-wall is seen to be so very thick, composed as it is of several lamina. Now the peculiar and interesting internal character alluded to is, that in the boundary walls of the cells or “joints” there occur always two op- posite terminal, and in some of them, in addition, two lateral, pairs of more or less wide (seemingly circular) pits—a character, indeed, found in other Rhodophycez, long ago pointed out by Nageli*; but what would appear to be novel (so far as I know, indeed, it is unprecedented in other departments of plants $) is, that each of these pits is covered by a comparatively large lid or stopper, not organically united or persistently adherent; at least very small accidental force seems to suffice to dislocate and remove them. In Mr. Moseley’s example, in every joint of the plant, whether of the main stems or rachises or of the ultimate ramifications (except the apical joints), there occurs a “ pit,” with its “ stopper," at each end of each joint—that is, at least two such in each cell (see the figures passim); in the apical joints there occurs but one (at the base). In many cells of the rachises and of certain lateral cells (presently to be noticed) there occur besides yet other pairs of smaller pits and stoppers, as will be hereafter explained. Why these pits * ‹ Die neuern Algensysteme.’ + Something optically resembling the “ stoppers” above referred to presents itself during the conjugated state of Mesocarpus nummuloides. It is well known that in Mesocarpus the spore is formed within the transverse canal (formed in the characteristic manner between the two conjugating cells) by the passage thereinto of their chlorophyll- plates enclosing the granular and oily contents, which become massed in the middle, whereupon ensues the division of the now H-shaped cell into three by formation of two septa, one to each side of the transverse canal, the middle cell becoming the spore, the two limbs (the parent cells) cut off seemingly as effete, though still containing colourless slightly granular plasma. As maturity advances, and when the spore presents its three consecutive walls, the median brownish and coarsely scrobiculate, it is seen to be ultimately freely suspended in the surrounding medium, but still posed between the somewhat geniculately bent pair of parent cells. Now, in addition to the original septum cutting off the primary parent cells, the outer apertures of the disjunct transverse canal, now separated at both sides from the free spore, is seen to be covered by a plano-convex “ lid,’ much thinned off at the periphery, of a semipellucid, somewhat “ pearly,” aspect, and sufficiently strongly calling to mind the appearance of the “ stoppers " in Ballia above referred to. This lid in Mesocarpus may be seemingly composed of a condensed kind of mucus. What its purport, I could not hazard a conjecture; but it certainly appears puzzling that so much pains, as it were, should be expended on hermetically * sealing up” these apertures in a structure whose function would appear to be over, containing, indeed, some as yet living (?) plasma, though robbed of its chlorophyll, but nevertheless seemingly destined to be but east off as of no further moment. The **lids" here adverted to are often, indeed, sufficiently obvious; yet I do not know that they have been noticed in algal works. Their possessing a certain amount of resemblance to the stoppers in Ballia renders an allusion to them here possibly not altogether out of place, though in the case of Ballia the stoppers are within the cell-cavity, : but external to the “ primordial utricle," whilst in Mesoezrpus the corresponding structures are (at least ultimately) external to the eell—that is, washed by the surrounding water: still their deposition must presumably have been carried on, at least begun, whilst the spore-cell remained organically united or in intimate apposition with the parent _ Whether, as suggested by Prof. Dyer, the “hemispherical depositions of cellulose” sometimes met with on the inner side of the pollen-cell wall, е. у. in Cucurbita реро (Sachs's * Text Book, р. 32, woodcut v), are to be considered analogous to the structure (“stoppers”) here adverted to is probably questionable, but is doubtless deserving of further investigation. MODE ОҒ GROWTH OF BALLIA CALLITRICHA. 213 are formed at all, or what special function they subserve, is not apparent; but the marvel is, that no sooner are they in existence than the openings seem to be thereupon again methodically plugged up. The terminal pits, with their corresponding stoppers belonging to the same cell, are as far as possible distant from each other, at the remote basal and upper ends of the cavity, whilst the pits appertaining to neighbouring cells (as is usual where pits occur in higher plants) are placed directly opposite to one another. Each pit is somewhat nar- rower at its inner end than at the mouth ; hence the two pits, viewed together, present an outline like that of a “pulley” in edge-view (Pl. XXVIII. figs. 11, 18). In the long and large cells of the main stems the pits are the largest, whilst they diminish gradually in size in the joints of the upper portions of the principal stems, or of the secondary or tertiary rachises, pari passu with their degree of tapering or attenuation (that is to say, with the decreasing size of the cells), until in the joints of the ultimate branches or pinne they become at last very minute, like mere granules or dots, but still quite per- ceptible (Pl. XXVIII. figs. 2 & 14). So far as can be learned by a close examination in joints which have not become injured, the primary membranes of the joint-cells seem to close up the bottoms of these adjacent pits; at least, a line can readily enough be discerned occupying that position (figs. 11, 16, 18). But frequently this appears, on examination with a high power, to have become obliterated ; not only so, but instances were noticeable where the contracted protoplasmie eontents of one of the joints (now quite colourless from the action of the spirit) had become from some cause thrust through the foramen (thus proving that these are real openings, not mere markings or even alterations of density or otherwise), the mass of contents (thus one half in one cell, the other in the other) roughly forming a figure-of-8 shape, owing to the constrietion caused by the narrowness of the aperture. Corresponding to this gradual diminution in the size of the pits of the stems and secondary branches onwards to the ultimate pinnules, there is not only a gradual diminu- tion in the size of the “stoppers,” as mentioned, but also a gradual modification of their figure. In those appertaining to the cells of the principal stems they are of a figure approximating to two thirds of a sphere; in the upper joints and in the smaller cells of the secondary and their lateral rachises they are of a figure nearly hemispherical (or kettledrum-shaped); and by degrees, as the cells diminish in size down to the smallest, they assume a nearly plano-convex figure, until finally, in the very smallest cells, they appear like mere dots, or brightish rounded or discoid granules. In all, it may be said, there is a more or less evident slightly projecting rim around the flat side, which rim often presents a slightly crenulated appearance (Pl. XXVIII. fig. 12). In those of the smaller dimensions and of a plano-convex figure, however, the margin appears to be somewhat “ bevelled off," and the crenulations forming an inner series. The convex or hemispherical surface is always quite smooth, and even glossy in appearance; ш dried examples it may become more or less broken or altered. The substance appears of a colourless, semitransparent consistence. It is the plane side which is apposed to the Opening of the pits, than which it is а good deal larger, closing it up like a lid—the SECOND SERIES.— BOTANY, VOL. I. ан 214 MR. W. ARCHER ON THE MINUTE STRUCTURE AND convex surface being towards the cavity of the cell. Strictly speaking, that portion of the superficies had better be called truncate rather than plane, as, especially in the larger examples of these bodies, it сап readily be seen to be hollowed out or ex- cavated in a rather irregular manner (Pl. XXVIII. figs. 11, 16, 18). Sometimes this hollowing out is a mere simple concavity ; or this may show at its middle а secondary depression, or even a tertiary somewhere about the middle of the secondary, with some- times a narrow inlet, then roundly expanding within. Thus the more or less irregular excavation may sometimes reach to a depth of a half to three quarters of the thickness of the whole body, but more frequently to only a quarter to a third the depth. Sometimes the excavation is nearly vertical laterally, and its bottom flat, in place of forming a concave depression, whilst that of an interior median secondary depression may be rounded or flat; and sometimes (though rarely) a narrow perforation may even pass right through to the hemispherical surface. In the larger examples especially, the inner surface of these inwardly variously shaped, though usually at their periphery pretty nearly circularly bounded excavations is seen to be not smooth but rough, even rugged, and of a somewhat granulated aspect. When the two knob-like covers or stoppers are undisturbed in situ—that is to say, placed directly opposite to one another, each upon the opening of its own pit, and thus closing it up—and when the two together are viewed in conjunction with the outline of the two pits forming, as it were, a canal slightly narrowing halfway from one knob to the other, the whole offers a very curious and singular resemblance to a “rivet,” as it were binding together the neighbouring joints; for the canal of the double pit is sufficiently deceptively like the shank of the quasi-rivet, the variously figured stoppers (as described) passing for its “ heads,” which clamp the structure on both sides. But when, as often happens, the stoppers become dislocated (see woodcut, р. 230), the quasi-stalk (that is the double pit) of course remains, dispelling an illusion of which the observer might at first glance be pardonably enough the victim. These stoppers did not offer any evident reaction in presence of the usual tests for starch and cellulose; they disappeared totally on the specimen being boiled in caustic potash. They are, so to speak, external to the so-called * primordial utricle," yet in intimate contact or union therewith. One could occasionally see the cell-contents still closely adherent to the convex surface of a stopper, partially contracted, and stretched into а conical shape, more and more expanding in width away from the stopper. Oftentimes the tension so induced caused the dislocation and pulling away of the stopper still coherent with the mass of plasma; or it might be “ let go” halfway ; at other times, and indeed more frequently, the cell-contents contracted, leaving the stopper i» situ. In otherwise seemingly empty cells the stoppers, if not in their places (see woodcut), as is very fre- quently the case, could be found lying about in any part of the cavity and in any post tion. By their comparative dimensions one might safely enough predicate the particular pits to which they doubtless had belonged, and whence they had become dislocated (Pl. XXVIIL fig. 3). — But on examining the organization of this pretty alga, and in trying to gain a know MODE OF GROWTH OF BALLIA CALLITRICHA. 215 ledge of the remarkable structure of which I have been endeavouring to convey an idea, one soon encounters an appearance, at first а very puzzling one, which, especially when viewed under à low power of the microscope, is readily misleading. Prof. Wright was the first to suggest the true solution of the contradictory appearance alluded to. But before I try to deseribe this, it will be necessary to convey an idea of the external or lateral form of the joints of the stems and rachises. The cells of the “© stems” and of each principal and secondary rachis of Mr. Moseley's plant are about + to $ as broad as long, and of a figure which has somewhat the aspect of an elongate oval (Pl. XXVIII. fig. 3). When more closely studied, especially under а higher power, they are seen to be of a cylindrical form for about $ or 3 their length from the base, at which point they become narrowed upwards, but in such a manner that thence to the summit the outline on each side gradually assumes a somewhat sigmoid eurve, the inward flexure at the upper portion being greater and somewhat more arched than that at the lower. Іп other words, at the place where the nearly cylindrical lower portion of the cell ceases it becomes gradually contracted; the outline, however, does not form any angle, but is rounded off and is thence continuously curved inwards, then upwards for a time almost vertically, soon imperceptibly beginning to be unin- terruptedly, but more rapidly, curved inwards on each side to the summit. Тһе diameter of the cells at the summit is usually about 2, or sometimes less, or it may be, on the other hand, nearly 1 of that of the cylindrical portion. I have called the lower portion of the eells nearly cylindrical; but they are not absolutely so, as they generally seem to be very gently and gradually concave on the lateral margins, the concavity being deepest about halfway from the base to the point where they become narrowed upwards. On the whole they mostly appear to be very slightly dilated from the concave middle to that point; that is to say, the widest part of the joint, by some very slight degree, is that just before where they begin to be narrowed by thesides curving inwards. So much for the contour (be it always understood of the cells of the principal stems and rachises) presented in the ordinary view, so far as relates to their external (that is lateral) boundary. I say, the ordinary view presented; for the habit of this plant being repeatedly distichous, the principal stems, branches, rachises, and their pinne or ultimate terminal offshoots, all lie in pretty nearly a common plane; hence a portion detached, however small, almost necessarily comes to be viewed from one or other of its plane aspects. The description given above of the lateral contour applies, it will be of course understood, to the joints so viewed. Now when they are examined in that position, the lower end of each joint of the Principal stems and rachises seems to be of a more or less conical figure, and broadly and bluntly rounded at the extremity, and as if projecting downwards into the joint immediately below ; the upper portion of the latter necessarily introverted to receive it (Pl. XXVIII. figs. 3, 10, 11, 14, 16; Pl. XXIX. figs. 15, 18). Not only has the structure the appearance described, but (if we are viewing this alga under a higher power) a slight alteration of focus further shows the base of the joints as broadly rounded off at a higher level, and as if the apex of the joint immediately below it were in close apposition | 2H 2 216 MR. W. ARCHER ON THE MINUTE STRUCTURE AND thereto. Ifthe plant be viewed under a low power, the two contours—the broadly rounded and the deep conical—come into view nearly simultaneously, and we have the curious and contradictory appearance of the joints having, as it were, ¿wo bottoms; and this is what gives rise to the puzzling appearance alluded to. Now the pits, with their accompanying remarkable knob-like stoppers, are seen to occur in the nearly transverse septum over the seeming conical base. If there were a conical base of the upper joint, passing down into and introverting the tip of the joint immediately below, how could the knob or stopper of the upper end of the lower joint get past the barrier, and become opposed, on its own pit, to that at the bottom of the upper joint? Without any opening or passage, mani- festly it could not be so deposited in its proper place. The curious appearance of the introversion of the upper end of the lower joint might, indeed, possibly meet its explanation by assuming an interspace, from this ordinary aspect of a bluntly triangular or conical figure, between the two joints; somewhat comparable to the very differently shaped interspace between the mutually introverted ends of several forms of Spirogyra, but in the present plant, indeed, only one (the lower) joint being introverted. But even if this were true, it would not explain the pair of directly opposed “ knobs ;" for one would be at the base of the upper joint, and the other must then be in an “ intercellular space," and therefore outside the cavity of the lower joint, thus offering really no explanation, and in truth presenting a puzzle just as dificult of solution as ever. The fact is that the apparent introversion of the upper end of the lower joint by the lower end of the upper joint projecting downwards into it, an interpretation which naturally occurs to every one on first looking at this plant, is illusory. So illusory is it, indeed, that Harvey himself, in describing his Ballia Brunonia, directly falls into the error, and quite misapprehends the true state of affairs *. He actually describes the structure, speaking of the joints of the rachis, thus :—“ Articulus singulus apice concavus, basi convexus, superior in inferiorem insertus, e cellula unica formatus, sacculam endochrome includens;" and afterwards he says, “ each joint is concave at its superior end, convex at its inferior, having thus a somewhat cordate figure, the convex end being inserted into the joint immediately below it, while the coneave receives in like manner the one above. These joints consist of a single cellule, and contain а bag of colouring matter which is collapsed in a dry state." Аз will be seen, the true structure is quite the reverse of Harvey's description ; whilst he is alto- gether silent as regards the pits and stoppers, whatever be their value or significance. Nor does it appear that subsequent authors had detected the fallacy or noticed these curious ** stoppers.” l The true explanation is, that the lower end of each joint is in fact deeply cleft or forked laterally—that is, forming two wedge-shaped subdivisions ; and these latter are tapered off towards each front, more or less conically, and then truncato-rotund, or rounded off subhemispherically. То some extent it might not be altogether inaptly compared to the top of a bishop's mitre. * Harvey, “ Description of Ballia, a new genus of Alge,” in Hooker’s Journal of Botany, 1840, p. 190. MODE OF GROWTH OF BALLIA CALLITRICHA. 217 Now, having before described the mode of narrowing upwards of the upper ends of the joints and their lateral contour (in “ ordinary ” view), it remains to say further that they are from both front aspects narrowed upwards—so as to offer a blunt wedge-shape, rounded at tip, if they were viewed laterally—and so as to fit into the deep similarly figured noteh of the lower end of the joint immediately above. Тһе boundary at the summit is somewhat coneave, adapted to the convex base of the interior of the notch of the joint above; but as it is considerably narrower, its concavity does not so readily strike the eye, and it might be regarded as straight or nearly so. Thus each joint might be said to be, as it were, equitant upon or to sit astride the wedge-shaped top of the joint immediately below ; its more or less tapering or narrower subdivisions (one of which is seen in either ordinary “ front view " of the frond) giving rise to the at first pardon- able enough error, that there is a projection into and introversion of the cell imme- diately below. Thus the true state of affairs is just the reverse of what one naturally is first inclined to imagine; it is the base (not the summit) of each joint which is introverted to make room for the tapering upper end of the joint immediately below it. But this introversion takes place in the peculiar manner and gives rise to the configuration described—this, as it were, ‘ splicing together ” of the joints doubtless adding to the tenacity of this rigid alga. That this is the true explanation I have satisfied myself by breaking up a small portion of the plant, and as muchas possible denuding little fragments of their branches ; and I thereby succeeded in getting a lateral view of a filament, otherwise necessarily, as has been mentioned, extremely difficult to obtain. In this view the marginal contour of a filament is pretty nearly straight from end to end (Pl. XXVIII. fig. 9). From the description already given it will be observed that—owing to the narrowing upwards with a curved outline of the upper ends of the joints of the principal stems and secondary rachises, and to the peculiar figure presented by the basal end of the joint above it, coming down to meet it with the curve appertaining to the inner angle of the great “notch” therein—opposite each “splice” there exists what may be called a recurring narrowing of the filament (Pl. XXVIII. figs. 3, 14; Pl. XXIX. figs. 13, 17). Owing to the peculiar contours of the joints being ignored in the figures extant, as a matter of course these shallow sinuses are not represented; the lateral margins are shown rather as those of broadly truncato-ovate cells. Still less is there any indication that, as is the case, each of these shallow sinuses is filled by a cell (Pl. XXVIII. figs. . 18,10, 11, 14; РІ. XXIX. figs. 17, 18). When fully formed, this intervening cell appears to be either quinquangular or sexangular, but in different parts of the plant of very varying proportional figure. And it is from this cell—never directly from the rachis- joint itself —that the lateral branches take origin. These lateral cells, which I would denominate *'ramification-cells " are either pen- tagonal or hexagonal; and the five or six sides (as the case may be) are of very unequal lengths. Two of the sides are apposed to the joints of the rachis; two beyond these are external and always unchanged ; and the one or two remaining sides intervene between it and a ramiñcation. The longest side in the pentagonal cells is that in appo- sition with the whole of the curved (that is, attenuated) upper portion of the joint of the 218 MR. W. ARCHER ON THE MINUTE STRUCTURE AND rachis. 'The next longest is that opposite thereto, external, and forming with the first an extremely acute angle (2); the third longest is that forming an obtuse angle with this latter ; and on it (3), as will be explained, stands the first joint of a branch; the fourth in length is that in apposition to the curved interior of the great notch in the lower part of the joint (4); and the shortest (which is very short) is that joining the last with the second longest (or that side bearing the branch) (5). When there are six sides the relative length of the sides is the same, and the additional side (6) is nearly equal in length to, and immediately under, that bearing the primary branch, and is itself (as will be ex- plained) surmounted by a minor shorter (stipule-like) branch. Now the mode of origin of these ramification-cells is seen to be by the formation of a peculiarly curved septum at one or (mostly) both sides of the joint of the rachis, previously cylindrical, or nearly so, all the way from its base to its summit—this septum beginning at its lower extremity from the point, about three fifths or two thirds from the base of the rachis-cell (before described), and ending at its upper extremity a little within the base of the joint immediately above (Pl. XXVIII. fig. 5). This septum (1) takes a sigmoid eurve as viewed from above, and is at the same time hollow or concave towards the ramification-cell, convex towards the rachis-cell, imparting the peculiar shape to the latter, as previously described. The cell thus eut off from the rachis-cell, as will be seen (it will of course be understood, when viewed from the “ front”), at first forms а scalene triangle—one side only, however (the outer), being quite straight, its lower end very attenuated, and its angle there very acute. In the formation of this septum we again encounter the curious character of an oppo- site pair of pits with their stoppers, every pit having its eorresponding opposite fellow, the likeness of а “rivet” being still carried out (omitted in the diagrammatic figure 5). The pair of pits in the new septum is placed about one third of the way upwards from the acutely angular bottom of the ramification-cell (а, fig. 10). Just as this cell is much smaller than the joint of the rachis, so the pits, with their stoppers, are much smaller than those at each end of thelatter. If ramification-cells are formed (as mostly, or indeed nearly always happens) at each side of the joint of the rachis (when formed at all), then the latter has two opposite lateral pits in its wall, in addition to the terminal; if only one ramification-cell be formed, then, of course, it has only one lateral pit. At this point, of course, the ramification-cell itself has but one pit—that is, the one opposite to and forming the fellow of the lateral pit of the rachis-joint. | The ramification-cell now soon grows out laterally in an inflated manner towards the upper end, the convex outgrowth being directed obliquely upwards (figs. 5 and 10). if viewed inverted, its noselike upper portion thus produced is now cut off by a curved con- cave septum, the concave surface being upwards—that is, the convex surface towards the cavity of the ramification-cell. The interposition of this septum makes the ramification- cell now five-sided—one of the sides (as before mentioned) being bounded by the septum just formed, the two adjoining sides, between which it intervenes, being the outer exposed portions of the ramification-cell now pushed out laterally, the upper one remain- ing very short and never further altered ; whilst the two remaining sides are, of course, those (as before mentioned) in apposition with the lower portion of the upper joint of the MODE OF GROWTH OF A2ALLIA CALLITRICHA. 219 raehis, the other in apposition with the longer portion of the lower joint of the rachis respectively. The upper (subhemispherieal) portion of the ramification-cell, thus eut off by the eurved, hollowed, oblique septum, is the first joint of a simple branch or (it may be eventually) of a subsidiary rachis (Pl. XXVIII. figs. 10 & 11). Before taking leave of this cell for the present, it is only necessary to say that at the middle of the septum is formed a pit—a double pit, it will be, of course, understood, each with its stopper. At this point, then, the five-sided ramification-cell has now, of course, two pits—one, as explained, opposite the corresponding pit of the rachis-cell, the other (younger) cor- responding to the pit of the basal cell of the branch; and as the former pair of pits, with their stoppers, are smaller than those at the ends of the rachis-cell, so the latter, in their turn, are smaller still than the former. À ramifieation-cell having reached this point may seemingly remain permanently so (in Mr. Moseley's plant many, by far the majority, do во); but certain of them proceed further. . If then they proceed to further development, the next stage is that from a point from about a third to scarcely half way from the outer side of the septum, cutting off the first branch-joint from the ramification-cell, a fresh curved and hollowed septum is produced ; this extends thence to about the middle of the lower and outer free and exposed (external) side of the ramification-cell, its concave side outward—that is, its convex side (as before) towards the cavity of the ramification-cell (6, in figs. 10 & 11). This latter has now reached the six-sided form above mentioned. Obviously the portion cut off therefrom is at first triangular, with only one exposed or external side—this being, in fact, the upper portion of the lower free and external side of the primary ramification-cell, which lower portion whereof still forms one of its boundaries, and of course still free and external. Again, we have to note the formation in this second septum of the opposite pairs of pits and their stoppers, occupying a position about one third of the distance from the lower end of the septum. The primary ramification-cell at this point, it will be seen, has now three pits—the third and last formed being that in the freshly formed septum, and placed at a point about a third of the distance from its lower end, and opposite to which it has its fellow pit and stopper. The triangular portion so cut off (which has one pit, with its stopper, in the septum between it and the other ramification-cell, being that first mentioned), like its prede- cessor, grows out at first laterally, and then obliquely upwards in a roundly inflated manner, this projecting portion being soon cut off by a fresh septum in such a way that | the lower remains as a five-sided cell. This latter may, in fact, be denominated a rami- fication-cell of the second degree (ғ 2, figs. 10 & 11); and the upper portion cut off therefrom forms the basal or primary joint of a secondary or accessory branch—the septum 30 cutting it off bearing again its medium pair of pits and stoppers, still smaller in size. _ The five sides of the ramification-cell of the second degree, when the basal cell of the accessory branch is cut off, are thus made up:—1st. The upper portion of what Was once the outer free side of the ramification-cell of the first degree, now pushed out laterally and obliquely (1”); 2nd. The septum cutting it off therefrom (2”); 3rd. The 220 МВ. W. ARCHER ON THE MINUTE STRUCTURE AND portion of the septum cutting off the first joint of the upper branch from the rami- fication-cell of the first degree (3”); 4th. The (whole of the) septum cutting off the first joint of the lower branch from the ramification-cell of the second degree itself (47); and, 5th, a very short side joining the two latter together, produced by the turning aside of the outgrowing extension of the ramification-cell of the second degree in its less advanced stage (5”). Leaving this lower minor or secondary branch (as in the case of the older and upper one) for the present, we come to a point whence it is far more difficult to follow out the structure. | On taking up the description of Ballia callitricha (Ag.), Harvey, as given by Agardh under the name Sphacelaria callitricha, the plant he had in view is thus, in part, de- scribed: ** Саше filis confervoideis vestito" *.... So also Prof. Harvey refers to the external ** plexus of filaments." : On looking at a piece of our plant, these “Ша confervoidea," or filaments, are indeed a prominent, but a very confusing feature, clothing, as they do, with a dense and intricately matted and tangled jungle the whole of the principal stems and some of the secondary rachises, the free ends of the filaments sticking about in every direction. This clothing might call to mind a tree densely covered with ivy, or, perhaps more fitly, might suggest a moss coated with Chroolepus aureum, only that the latter plant, when investing the moss, makes a smooth pile. The authors referred to seem not to have investigated the relations of this plexus of “confervoid filaments" to the more handsome distichously branched Badlia-plant. I therefore desired the more to try to **make head or tail" of the confusing object before me, so far as regards the accessory clothing of what might be regarded as the Ballia proper (Pl. XXVIII. figs. 1, 4, 15; Pl. XXIX. fig. 14). Not to speak of the general appearance and the similar thick-walled character of the | cells of these filaments, two characters would set aside any possible supposition that these were not really part and parcel of the make-up of the Ballia. One of these was the presence in every one of the here directly transverse septa, separating their compara- tively short quadrate joints, of a minute pair of pits with their stoppers (Pl. XXVIII. figs. 6, 7); the other, that when they branched it was seemingly by the interposition of a ramification-cell, but not of the same figure as those of the distichous stems and branches as described (figs. 6, 7). Looking more closely into the build-up of this clothing of the stems and some of the rachises of the plants, it is readily enough seen that this is most dense at the lower por- tions, less and less so towards the upper portion of the plant, until it at last is wanting. It is also readily enough seen, by focusing “through and through,” that at the lower portions the main stems are covered all round by a single layer of thick-walled elongate cells, mutually closely apposed, of various irregularly curved and crooked figure, or етеп somewhat rugged within—and that these cells are separated by oblique septa, these latter furnished with their median minute pairs of pits and stoppers, sometimes an extremity of some of the cells growing on so as somewhat to overlap the adjoining ж «Тсопев Algarum,’ no. vi. (t. 6). MODE OF GROWTH ОЕ BALLIA CALLITRICHA. 221 one, their apices rounded. Thus this cortical stratum of cells somewhat resembles a kind of pleurenchymatous or prosenchymatous tissue. As our examination is carried from below upwards, this cortical: stratum becomes more confined to the margins, leaving more or less of the stem-joints bare at the middle, until by-and-by it dis- appears altogether. Here and there, with patience, one can probably meet with a profile view of the point whence the cells of this pleurenchymatoid cortical stratum is given off from one of the multitudinous branches which stick about most confusedly in all directions. In Mr. Moseley’s plant these seem scarcely ever themselves to become branched, but form only indefinitely long “© confervoid” filaments. Whilst, as mentioned, the septa of the vertical cortical cells are oblique, those of the branches given off therefrom are ever transverse; but, like those of the former, every septum here, again, has its minute pair of pits and stoppers. Only occasionally one can get a satisfactory profile view of the place of ramification; and it can be seen that this process here too is sometimes brought about by the intervention of a “ramification-cell.”” But this does not maintain the same figure as that of a rachis-cell as above described. Here, however, as there, an oblique, sometimes curved, septum is formed just under one of the ordinary oblique septa of a cortical cell, reaching from a point, say, about a quarter of the length of the septum from the outer wall to a short distance down the lateral margin of the cortical cell. Thus a minor cell is cut off, forming in this view a scalene triangle, two of its angles being acute, one obtuse. Soon the upper portion of its free side (being in _ fact a portion of the original outer margin of the cortical cell itself) grows out laterally, and, by apical growth and repeated subdivision of its apical cell, gives rise to one of the long accessory filaments, every septum, as mentioned, having its pair of minute pits and stoppers. The cells gradually diminish in size ; in other words, the filaments gradually taper. The apical cell, when full-grown, is conical, its apex acuminate; but until then the apical cell is bluntly rounded. Whether, on the occasion of the giving-off of these lateral branches а ramification-cell is always formed, ог whether they sometimes branch off without that preliminary by a simple outgrowth, merely standing off, as they some- times seem to do, like a thumb to a glove, I could not decide. So much for the ramifications of the cortical cells and the mode in which they them- selves coat the principal rachises. Within one of these, so coated, one can see, by focusing down, shimmering through the investing cortical layer, the great pits and stoppers of the largest size, as well as the contours of the joints, with their ramification- cells giving off subsidiary rachises. "These latter project through the cortical stratum of cells; and by far the greater number of these subsidiary branches themselves are naked. At a first glance, the question might propound itself, why do the investing fibres “тап up" only a few of the subsidiary branches, and not all? why are they so much more dense low down, as compared with the upper portions of the stems and such rachises as they do invest? This leads to the inquiry, What is the origin of the cortical cells (we have seen the origin from them of their lateral branches), and what their relation to what may, as before, be called the “ Ballia proper"? And this is just an inquiry to which it is very .. SECOND SERIES.— BOTANY, VOL. I. 21 222 MR. W. ARCHER ON THE MINUTE STRUCTURE AND hard to obtain a reply; nor do I think that, without the opportunity to examine a Ballia-specimen kindly sent me by Professor Thiselton Dyer (also from Kerguelen’s Land, collected by the Rev. A. E. Eaton) (Pl. XXVIII. figs. 13-18), I should have been able to satisfy myself on this point from the specimen in Mr. Moseley's collection, owing to the extremely crowded and nearly impenetrably dense investing envelope. To try to follow out, then, the relation of the cortical cells to the rachises they invest, we must now revert to the ramification-cells of the second degree before described. These we left as so many five-sided cells—that is to say, after the formation of the septum cutting off therefrom the portion destined to become the first joint of the lower minor accessory branch. In certain of these yet another curved septum makes its appearance: this, as might be expected, from the later period of its formation, is, by comparison, thin ; and I have not been able to see in it the usual pit (or pair of pits) ; but I would not say that they may not be there (РІ. XXVIII. fig. 11, г). This septum reaches from a point about one third of the distance from below upwards of the septum cutting off the first cell of the accessory branch, to very near the lower angle of the cell (ramification-cell of the second degree). It is much curved, and has its convex side upwards or towards the cavity of the cell, and its concave side towards the portion cut off. 'The upper portion is, as will be seen, still five-sided, the new side (made by the new septum) being considerably arched. This portion seems never again to divide or grow in any way, thus as if of no further moment (ғ c^, fig. 18). The lower portion сиё off therefrom is at first subtriangular or subquadrangular, one side exposed—that is, external (being the lower outer one of the original ramification-cell of the second.degree, as previously described) ; the other a portion (scarcely one third) of the septum, cutting off the first joint of the accessory branch; and another the new curved septum—which latter, if it gradually passes as far as the lower angle, causes the new cell to be trilateral; if, however, it terminates (ere it reaches that) somewhere near the angle, and against the adjacent side of the ramification-cell of the first degree, then we have a cell of quadrilateral outline. "This last cell then, as will be seen, when it is produced, might possibly correctly be called a ramification-cell of the ¢hird degree (пе, fig. 18). The next stage seems now to be, that the outer lower margin of this ramification- cell of the “third degree” becomes, as it were, somewhat abruptly pushed or bulged out downwards at about its middle point—this bulge soon, however, taking a more outward and lateral direction, the thick membrane becoming somewhat doubled upon itself Sometimes one might fancifully suppose that this doubling outwards laterally got some assistance by a mechanical “dint” from without (т, fig. 17), imparted a little to one side of the nearly similar bulge from within ; for the outline now shows this twice bent appearance. This doubling upon itself, backwards and outwards, becomes more pronounced ; the lower portion becomes enlarged and inflated ; and soon there is presented a large ovate external addition to this ramification-cell of the third degree (cort. cell 1, figs. 17, 18). This accession, in length downwards (alongside of the rachis-cell below), P about equal to that of the older portion above when first formed, and in width twice as great as that of the original portion; at the same time this additional ovate external MODE ОЕ GROWTH ОЕ BALLIA CALLITRICHA. - 228 . portion is now itself about twice as long downwards as broad. This point attained, now another septum is formed, of peculiar curvature, and stretching into the cavity of the large ovate new extension thereto, and thus, whilst it cuts off the larger portion, still leaves the ramification-cell of the third degree larger than it was originally—in fact, nearly twice the size. Having progressed so far in the examination, the structure becomes more and more difficult to follow. However, the septum last mentioned seems to take a curve from a point opposite to the sudden outer doubling back of the wall, and with a considerable sweep (something like the curved hook of a “ note of interrogation ”’) then passes quickly round to a point pretty nearly opposite the lower angle of the rami- fication-cell of the first degree. Thus the ramification-cell of the third degree, viewed in connexion with the portion added to it, has acquired a somewhat sigmoid curvature, the extremities broadly rounded ; or it presents the figure of a subcylindrical cell with bluntly rounded apices, somewhat suddenly twice bent near the middle. Now, in the lower portion сиё off (after forming this curved cell, and not belonging thereto) we have the first cell of a dependent appressed filament, thereafter elongating and repeatedly subdividing by oblique septa, and contributing to form the cortical stratum described. The cells go on so growing until they reach the node below, and there seem to divaricate. But the bent cell above seems to have more to do: it sends a protuberance out from each side (that is, towards each “ front ” of the flat frond of the Ballia plant), each of these forming a basal joint of further dependent cortical filaments, which, by passing downwards somewhat obliquely, cover more of the front of such main stems and principal rachises as proceed to this development (cort. cell 2, figs. 17, 18). At what point the very first septum is formed I could not make out exactly. Arguing from the analogy afforded by the mode of branching lower down, Г suppose a septum is very early interposed upon the formation _ of a convex expansion, at both “ fronts ” (of the frond) of the lower portion of the bent cell—such septum eutting off that portion as the first joint of the cortical filament, which is dependent at each side of that previously described, these two having the latter between them, one running down on one front of the frond, the other on the other aspect. It seems, however, sometimes to happen that no septum is formed in the ramification- cell of the second degree, and thus no portion thereof is cut off behind as useless; but the process described can go on, this omission notwithstanding. There is then no ramifica- tion-cell of the “ third ” degree formed. That of the second degree having had cut off from it the first cell of the lower accessory branch, proceeds to develop from its basal portion, by the outgrowing thereof and formation of the curved septum as described. The extreme lateral cortical filament, and at both front surfaces (of the frond), gives off the protuberances (keeping the former between them as described), which grow into further cortical filaments, directed more across the front aspects of the rachis as described. The mode of branching of the dependent cortical cells is by repeated dichotomy. Close to the rounded apex of the terminal cell two septa are formed (making a У); these оз course cut off two chords or cells of at first plano-convex figure, which, expanding and ` elongating downwards, grow side by side as the first joints of two collateral branches. As the filaments grow downwards, occasionally thus branching, they cover first the margins, then more and more of the frontal surfaces of the stem; and they keep closely 212 . 224 MR. W. ARCHER ON THE MINUTE STRUCTURE AND apposed to one another and to the cells of the stem or rachis, leaving (especially lower down) very few “ intercellular spaces " (that is, intermediate bare places of the stem). When these occur they are of irregularly acutely lanceolate outline. Тһе largest inter- cellular зрасез are those, of course, through which the lateral branches or subsidiary rachises project—left necessarily by this descendent investing cortical stratum of cells as they pass downwards to each recurrent node or joint of the great stem or rachis, and from each of which in the same way new ones are given off in their turn. If a lateral rachis possesses such a cortical stratum of cells, they belong to and emanate from its own nodes; for such dependent cells on arriving at a branch never become directed upwards to cover it, but the system passes downwards by the same stem to the base. The cells forming this cortical stratum soon begin to give off their own branches (as described), and more and more copiously as they descend, until the great confused tangled ** plexus " is produced. . In order to complete the description of Mr. Moseley's plant we must now briefly revert to the ramification. We left the basal joint of a principal branch just formed by the cutting-off of the same from the ramification-cell of the first degree; and, in the same way, we left the first joint of a minor or aecessory branch just formed by the cutting-off of the same from the ramification-cell of the second. degree; but not all, indeed com- paratively only a few, of the ramification-cells proceed to this latter development. The greater number of the ramification-cells remain as of the * first degree "—that is, Вуе- sided; and the cell cut off gives rise, by continuous apical elongation and repeated . division, to a more or less long (sometimes very long), often very stout, filament. There is in such cases no septum producing a ramification-cell of the second degree formed ; and therefore no basal or accessory branch originates (Pl. XXVIII. fig. 3, its upper part). Each branch so emanating from a principal stem has (it need hardly be said) nearly always a similar one opposite thereto, arising, as described, from its own ramification- cell; and from every node for long distances such a pair of branches may be thus distichously given off. But it sometimes happens that a branch may be without an opposite corresponding one: the lateral margin of the side so deficient remains straight from bottom to top ; no ramification-cell is cut off—no branch (parts of fig. 2). And just in the same way in Mr. Moseley's plant these lateral branches (pinne) may be, and often are, quite simple for the whole of their length (that is, with linear sides), no ramification-cells being cut off, therefore no pinnules formed. But the overlapping of the laterally subdivided bases of the cells towards both fronts of the frond is more or less carried out in the joints of the lateral branches. The base of the first cell of a branch arising as described, and of each succeeding one, is, like that of the rachis-cells, laterally notched, but comparatively not at all so deeply: thus, like them, each cell sits astride the upper portion of the cell below; that of the first cell so bestrides, of course, the adjacent side of the ramification-cell. Тһе basal subdivisions of the cells of the branches which partially overlap the cells below, one on one front, the other on the other front of the frond, however, are not so prolonged and so conically tapered off downwards as are the cells of the principal stems or rachises, but are rounded off subsemicircularly, or with a merely convex outline. Still those nearer the base of E | | MODE OF GROWTH OF BALLIA CALLITRICHA. 225 the branch or lateral rachis are more deeply overlapped (in the manner described) than are the upper cells. Ав one proceeds upwards the overlapping is seen to be less and less pronounced; that is, the curve downwards in front of the subdivision of the bases of the component cells of the branch becomes less and less arched ; whilst near to or at the top they appear hardly at all to overlap. The cells very gradually diminish in size upwards ; in other words, the branches gra- dually taper. Тһе increase in length of the branch is effected only by the growth and sub- division of the terminal cell, the next lower cell never subdividing again (unless, of course, it should produce ramification-cells). During the progress of the growth of the branch the terminal cell appears to be rounded at the top ; but when the branch attains its full length, and apical growth is over, the terminal cell becomes drawn out, long and conical, its summit acuminate, and its membrane thickened. Imay add that the cell-membrane, especially of the principal stems or ramifications, appears to be irregularly and longitudinally distinetly, though faintly, striate. With the exception of the **stipule-like" secondary branchlet (given off from the ramification-cell of the second degree), the ршиз (in Mr. Moseley’s plant), in com- parison with other forms, very seldom ramify ; that is, they give off pinnules only here and there very fitfully. However, this lower external branchlet, unlike an ordinary pinna or pinnule, as has been seen, is not an ultimate emanation from the same series of cells (or, in other words, from the same subsidiary ramification) as the pinnæ proper, but it is a collateral and ultimate emanation from a series of cells (forming a main stem or а ramification) of a generation back. This collateral or accessory ramification, then, is as much an independent branch as the older one above, although it might (in error) be taken for a pinna springing therefrom. But inasmuch as this accessory branch is but a subsequent emanation, primarily owing its origin to the ramification-cell of the second degree, which in turn was formed by the subdivision of the primary ramification-cell, it could not therefore exist without having been preceded in time by the (larger) branch immediately over it, at the outside base of which it stands. : The secondary or subsequent rachises in Mr. Moseley's plant, and their simple branches or pinnze, are usually pretty stout, taper very slowly, and stand off from one another at an acute angle, here and there often in a very scattered and irregular manner, giving off solitary branchlets, or little groups of pinnules—it need not be said, each ever preceded by a ramification-cell. It will be seen, then, that we have a plant of a соп- siderably long-drawn-out and fastigiate habit. Though the branches and pinne lie pretty much in a common plane throughout, still the plane may be more or less altered through the extent of the aggregate formed by a plant of, say, a few inches in length (PL XXVIII. fig. 2). It would be a very diffieult matter, perhaps hardly possible, to trace out correctly the different layers or laminze of the very thick walls of the rachises up into their rami- fications. It seems that an outer layer of possibly numerous thin laminz, forming a portion (the upper) of the original wall of each joint of a stem or rachis, reaching from ow upwards, becomes diverted laterally, and thus continued forms the outer layer of the lower exterior wall of the ramification-cell. But soon afterwards, owing to the 226 МВ. W. ARCHER ОМ THE MINUTE STRUCTURE AND tension to which it has been subjected during the outgrowth of the latter and of the ramification seated on it, it becomes gradually “thinned off to nothing” and, excepting any little of it left higher up (the mere softened or macerated remains), totally disappears. Its seeming (and why I say “seeming” опу, will appear below) total disappearance is hastened when a ramification-cell of the second degree is formed, owing to the more sudden lateral diversion and increased tension of this layer caused thereby. The inner stratum of the free side of the ramification-cell of the first degree is deposited simul- taneously with, inasmuch as it is a continuation of, that septum, shutting off therefrom the ramification-cell of the second degree; and it is carried onwards round the interior of the ramification-cell of the first degree. The lower portions of this layer converge towards the acute base of the cell, and there, mutually apposed, seem to pass vertically downwards inavery compressed state ; moreover as a single layer between the outer layer of the base of the stem- or rachis-joint (from which we started) and an internal layer, which passes round the interior of the rachis-joint and at its upward parts (on both sides usually), it cuts off therefrom the ramification-cells of the first degree. Now when а secondary branch has fully grown, if we carry our examination upwards along the exterior ofthe primary branch (secondary rachis) immediately over the former, we see an outer layer somewhat sharply marked off from the inner, and this likewise * thinned off to nothing" upwards. Now I believe this layer was once in continuation with that below the subsidiary * stipule-like " branch, and whieh latter, as we saw, is (and of course always was) in continuation with the lower external layer of the stem- or rachis-joint itself; that is to say, it had been in continuation therewith prior to its having been forced asunder by the continued stretching and tension caused by the growth in the first place of the ramification-cell of the second degree, and afterwards of the branchlet seated thereon. Further, this upper outer layer last alluded to must have itself once formed part and parcel of the upper part of the original wall of the stem- or rachis-joint, anterior to the formation of the septum cutting off therefrom а ramification-cell of even the first degree—from whose expansion laterally it got its first impetus outwards, far away as it has now been transported from the place where it was originally deposited. And this is why I spoke of its seeming disappearance. Suppose now that the basal joint of à branch goes on to produce a ramification-cell (and a branchlet), we find immediately over the same a piece of external layer “ thinning off to nothing” as before; this latter, again, must have been, I presume, pushed up and out from the ramification-cell just below it, and previous to that have formed so much of the upper portion of the basal joint of the branch, and so on. Possibly it may be assumed that something similar occurs, were it traceable, in ordinary branching confervoids ; but so common is a gelatinous softening of the outer layers in such, that any traces of the mutual superposition of lamin:e get lost. ; On comparing this plant with examples of Ballia callitricha from the University herbarium, kindly placed at my disposal for comparison by Dr. Perceval Wright, one isat once struck by the very great difference in habit between the two. Indeed the same observation holds good as regards the further forms or species В. Hombroniana and B. Brunonia, distinguished by some from, and by others combined with В. callitricha. ылады EET РЕТ GEENE EEEE MODE OF GROWTH OF ZALLIA CALLITRICHA. 227 I leave the thoroughly distinct forms of В. Robertiana (Harvey) and В. Mariana (Harvey) out of view, not having investigated them with sufficient pains to be able to contrast them in their relative bearing to the points presented by Mr. Moseley's plant; but confining our attention to those forms combined under the designation В. calli- tricha, we see they are all of beautifully more or less well-developed plumose habit, very unlike Mr. Moseley's form. It would, however, be trenching on rather unsafe ground for me to venture to distinguish this plant as a new species. In B. catlitricha and the forms conjoined with it under that designation, the whole that has been said in relation to Mr. Moseley's form, as regards the pits and stoppers, the ramifieation-cells and mode of evolution of the branches and cortex, in a general way applies. The differences in the different forms lie in the various figures of the different constituent cells forming the build-up of the plants, until we come to the ultimate ріппге with their pinnules. Unlike Mr. Moseley's plant, where these ultimate pinnules are rarely and only fitfully given off, in the В. callitricha forms they are methodically and pretty uniformly given off from the cells of the ultimate little rachis—always by interposition of a ramification-cell of the first degree only, and so as to produce à plume of ovoid or lanceolate or hastate outline, according to the ** species." It becomes worthy of remark that whatever the outline of these plumes, and conse- quent effect or habit of the plant in its entirety, the lowest outer component thereof, looking like a pinnule of the same nature and origin as the rest, is not really so, but is in fact the homologue of the lateral “ stipule-like” secondary branch described in Mr. Moseley's plant :—there mostly, or at least often, solitary ; here, along with the close array of pinnules, contributing to make up the ** plume," and rendering itself quite conform- able with the true pinnules. That is to say, it is a true secondary branch, being ап emanation from the ramification-cell of the second degree belonging to the older rachis from which that bearing the pinnze starts, as this latter is an emanation from the preceding ramifieation-cell of the first degree; and from the ultimate minor rachis standing thereon, in turn there emanate all the pinnules standing above this secondary branch, which is conformable with the piunules, and forms part and parcel of the ** plume." It sometimes, indeed, happens that this lower branch grows inordinately in length, refusing to comport itself like the line of pinnules above it; I did not, however, ever notice a case in which it branched or gave off secondary pinnules or “ pinnelle.” In Mr. Eaton’s form, sent by Prof. Dyer, the inordinate growth of this accessory branch seemed more common. The ultimate terminal or apical **plume" of each ultimate rachis in the typical B. callitricha must, however, be excepted; for its pinnz аге wholly made up of short stout tapering branchlets, each standing on a ramification-cell, all of the first degree, а ` continuation of the penultimate rachis being its rachis. In other words, these do not go on to become ultimate rachises themselves, do not produce ramification-cells, hence no “pinnelle ;" but together they form an apical “ plume," of more or less lozenge-shaped general outline; and sometimes the upper cells remain inactive, produce no ramification- cells, and do not ramify. Sometimes, indeed, one or two of these apical pinne (branchlets) may make an attempt from a few of their cells to give off ultimate pinnelle, 228 MR. W. ARCHER ON THE MINUTE STRUCTURE AND though these sometimes reach only the formation of a ramification-cell—as indeed the effort to produce pinnelle at the apices of some of the plumes lower down may get no further than the same preliminary stage. In the “ typical" form the ramifications given off from the cortical stratum are much more prone themselves to branch than in Mr. Moseley's. Several branches (from three to ten) are given off, generally alternately, in quick succession, and radiate in a kind of fan-like manner; each possesses always a preceding ramification-cell (five-sided when the branch is formed), as in Mr. Moseley's plant. Тһе pits and stoppers are present, but not so conspicuously evident. The differences between the “ typical" В. callitricha and В. Hombroniana are very graphically shown in Mr. Tuffen West’s delicately rendered drawings. More distinct still than they are from each other, appears Mr. Moseley’s plant from either. At first glance, Ballia scoparia, Harvey, seems hardly to belong to the same genus. Distinct as is the habit, and divergent as is this plant in several details from the pre- ceding forms, there is still a good deal in common; but possibly, after all, the same characteristics may yet be found to be more forcibly presented in other algee. So far as regards the pits, they are not confined to Ballia in Rhodophyces ; these, as “ pores,” have long ago been pointed out by Nägeli. It is the “ stoppers” (so called) which I fancy to be new. | Now in B. scoparia the pits are generally pretty evident, the stoppers not readily seen, presenting themselves as only little thin discoid bodies (Pl. XXIX. figs. 8-9). It likewise possesses a plexus of threads on the principal stems, of which more below. . Like the former, this species branches by interposition of *ramification-cells;" but these and the mode of branching are quite dissimilar. Its general habit is aptly enough indicated by Dr. Harvey's specific name :—long-drawn- out and fastigiate; the “branches” scarcely thinner than the “stems,” most irregularly given off at no particular intervals, and only at one side, and branches fitfully again branching (Pl. XXIX. fig. 2). The branches stand off with a slight curve from the cell on which they originate, and then proceed upwards at an acute angle to the “ stem.” The ends of the joints appear to be slightly hollowed; and the adjacent upper and lower ends mutually somewhat overlap, the outer lower margins being convexly rounded (figs. 8, 9). The upper portions of the branches give off unilaterally a succession of 8 few (1-5-6) of short spine-like uncinate bluntly acuminate branchlets (fig. 3); and when these are formed, that branch which bears them grows upwards no more; that is, its terminal cell ceases to subdivide, and its apex becomes acuminate, like the lateral spine- like branchlets. Now the formation of the ramification-cell is puzzling. When a young one is viewed edgeways, it seems to form a lenticular hollowing in the thick wall of the joint (fig. 6); but how that could so originate is very questionable. It may be formed by a deposition of membrane all round the joint, but so as to shut out a portion of the ** contents," which assumes that figure. Viewed from the “ front” (fig. 9), this as yet very thin cell is seen to be of broadly elliptic figure, showing a minute sharply marked pit and stopper exactly in the centre. More grown, and viewed edgeways, one of these cells appears E í— —OOEÓ QR MODE OF GROWTH ОЕ BALLIA CALLITRICHA AND В. SCOPARIA. 229 much inflated inwardly towards the cavity of the joint, the median pit at least readily enough made out (Pl. XXIX. fig. 5). It seems sufficiently surprising the comparatively considerable number of cases in which no further development appears to ensue. Sometimes one can trace three or four of these ramification-cells in as many joints in immediate succession at various parts of a “ tuft” of this alga, and frequently in other places scattered here and there a single one. Possibly the rate of growth may be but very slow; and many of those incipient ramification-cells apparent іп a given example probably would have gone оп, were the plant undisturbed in its native habitat, to the development at some later period of further ramifications. This process seems to take place simply by the projection of the outer side of the ramification-cell in an inflated manner (fig. 5), gradually expanding in an upward direc- tion. When about half the length of an ordinary joint is attained, a septum is formed, the apex grows onwards, then a fresh septum, and so on, until the definite length of the branch is attained; the upper joints finally give off the acuminate spine-like branches, and the apical cell becomes itself almost spine-like and acuminate. From the peculiar figure, as described, of the ramification-cell the branch has somewhat the appearance as if it issued through a lateral cleft of the wall of the joint bearing it; and the original inner eonvex boundary of the ramification-cell, being wider than the portion issuing from the side of the joint in which it is seated, lends to the whole much of the appearance of being a kind of graft, as it were, let in at that side of the joint whence it emanates (fig. 8). Notwithstanding that the mode of origin of the ramification-cells here seems to differ from that in the cortical cells of Mr. Moseley's plant, still, when the branching is effected, there is a sufficiently striking similarity in appearance in the “ отай-Нке” aspect of the insertion of the ramification, and in the mode in which it stands off from the joint bearing it (Pl. XXVIII. fig. 7). Towards the lower end of the plant decurrent branches are given off somewhat similar to those in the callitricha group; but here they do not seem to form a close “ cortex." Their mode of origin seems to be very similar to that of the ordinary branches; but I was unable to alight on an example i in as early a condition; primarily the ramification- cell seems to be similar. : The ramification-cell here, however, when the decurrent branch is formed, does not seem to offer as much of the appearance of a “ graft" let in at the side as does that of an ordinary branch (Pl. XXIX. figs. 10, 11). When it is fully formed it appears of a broadly ovate figure truncated below—that is, cut off there by the septum, dividing from it the next joint of the dependent branch; its inner lateral margin projects in a convex manner towards the cavity of the joint whence it emanates; and its outer boundary projects in a convex manner towards the surrounding medium. But this cellis always Subdivided by a septum (2, in figs. 10, 11) into two unequal portions, the origin or Significance of which I have been unable to make out. This septum runs in a curved manner, its convex side inwards, from about the upper apex (the whole plant being viewed vertically) of this truncato-ovate cell to near the outer lower angle (formed by the outer SECOND SERIES.— BOTANY, VOL. I. Әк 280 MR. W. ARCHER ON THE MINUTE STRUCTURE AND bounding wall and the first transverse septum alluded to); it thus cuts off а portion (the outer and smaller) of doubly convex figure (? in figs. 10, 11) from the other portion (the inner and larger) of triangular figure, bounded, however, by two curved lines and one straight (7 c in figs. 10, 11). Тһе septum, concave towards the latter, and separating it from the original joint, possesses its minute pair of pits and stoppers; so does the flat septum separating off the first joint of the decurrent branch, as well as the subsequent septa; but there seems to be none in that septum, convex towards this triangular cell, which cuts off the lenticular portion alluded to. At what period this latter peculiar septum is formed, or for what purpose, as I have said, I could not make out; and it remains to me one of the puzzles connected with this plant. It seems, as it were, like a ** ramification-cell " which never, however, goes on to produce a branch. The apical cell, whose punctum vegetationis has impressed upon it the downward impetus, continues to grow onwards, repeatedly dividing by a transverse septum. It is not long ere this dependent branch begins itself to ramify ; perhaps even the very first- formed joint may produce а ramification-cell quite like those of the ordinary erect branches ; and, as in their case, often ** ramification-cells " are present that do not seem to go any further. When ramifications are formed, which, indeed, is still very frequent, they stand off more approaching a right angle than do the ordinary branches (see Pl XXIX. fig. 4); and they have the same “ graft-like" appearance; the subsequent septa are transverse, and the pits often very obscure. Towards the base of the plant ultimately a dense irregular plexus of branching filaments is in this manner produced. Drawing, in outline, of a joint of the stem from a portion of Mr. Moseley's example of Ballia callitricha. It shows the junction. and: eoadaptation of the two cells, their partial invagination and the seeming double bottom, the double pit, and the pair of stoppers thrown out of position. One of the latter appears in profile; the other exhibits its truncate side or that surface which is apposed to the opening of the pit. Magnified about 400 diameters. MODE ОЕ GROWTH ОЕ BALLIA CALLITRICHA AND В. SCOPARIA, 231 DESCRIPTION OF THE PLATES. Puate XXVIII. Illustrations of the minute structure and mode of growth of Ballia callitricha. Figs 1-12 are from the fragments of this alga, gathered by Mr. Moseley, of H.M.S. ‘Challenger,’ in Kerguelen’s Land. Figs. 13-18 are from those obtained in the same island by the Rev. A. E. Eaton, during the Transit-of- Venus Expedition, and placed at the author’s disposal by Mr. W. T. Thiselton Dyer. Fig. 1. A portion of Mr. Moseley's plant : natural size. : Fig. 2. A portion of same, showing its habit and the fitful manner in which the ultimate branchlets are given off, and the paucity of any thing like the regular “ plumes ” of the other forms: x 20. Fig. З. A smaller portion, more highly magnified. The lower part of the figure shows the 6-sided ramification-cells, which therefore bear, besides the ordinary pinnz (branches), the accessory minor branchlets. Тһе upper part of the figure shows the 5-sided ramification-cells, which therefore are without the accessory branchlets. In some of the joints of the rachis the stoppers have become dislocated, and lie apart within the cavity. х 50. Fig. 4. A node or joint of a rachis, showing the dependent cortical investment and the bases, with the intervening “ ramification-cells," of four simple branches (i.e. without the accessory ones), the two lower being, of course, at the top of the next abutting joint beneath; the opposed pairs of stoppers can be seen above and below, the stopper at the top of the joint having become slightly displaced. x 100. | Fig. 5. Diagrammatic figure (founded, however, upon a young growing branch actually observed), showing the youngest and more advanced conditions of the ramification-cells. Тһе letter ? is placed against the septum in each case produced to cut off the triangular portion of the rachis-joint to become the ramification-cell. Тһе letters r c indicate in all cases the “ ramification-cell." The growing apical cell is still rounded ; when fully formed it becomes acuminate. Fig. 6. A branch given off from a cortical cell—that is to say, one of the component filaments of the great . plexus standing off from a main stem or principal rachis: greatly magnified. Fig. 7. Another filament giving off a branch in two places, also highly magnified : 7 с, ramification-cell. Figs. 8, 9. Side views of a node of a rachis under a still higher power. Fig. 10. Diagrammatic figure of the mode of growth of the ramification-cells, founded. on. cases actually observed in specimens: r c!, ramification-cell of the “ first” degree; r с’, of the “ second” degree. The figures 1-6 represent the different sides of the cell as adverted. to in the text ; aa, lateral pits and stoppers of the joint-cell ; 4, the terminal pits and stoppers. Fig. 11. Similar but more advanced, showing now the base of the accessory branch (hence necessarily the ramification-cell of the second degree) fully formed. The figures 17-8! represent the different sides of this cell, as adverted to in the text (the figures 1-6 and the letters а and 0 as in preceding). These two figures indicate the manner in which the projections from the base of each of the joints overlap and sit astride upon the apices-of the joints below, thus giving rise to the old error that they passed down into a pocket-like introversion of the same. Fig. 12. A stopper of large size, seen from its flat surface. Fig. 18. The Rev. A. E. Eaton’s plant, of natural size. Fig. 14. Same magnified, showing habit; near the top the branches simple, lower down showing 232 MR. W. ARCHER ON BALLIA CALLITRICHA ЕТС. accessory branchlets (the ramification-cells with the change mentioned in text), lower down still the branches showing pinne. x 50. Fig. 15. A portion of same plant more magnified. Тһе branches have (all but one, which remains simple, having formed no ramification-cells) become secondary rachises ; and the ramification- cells of the principal rachis have given off cortical cells : these have not yet completely clothed the rachis, in fact only on the sides; nor have they given off any lateral branches. In one place (fitfully enough) the ramification-cells on each side of a secondary rachis have given off cortical cells. The large pits and stoppers of the median rachis are very prominent; they are much smaller in the secondary rachises, gradually diminishing in size upwards, smaller still and gradually diminishing in the collateral branchlets and pinne, but, though reduced to mere dots, are always present. х 50. Fig. 16. A small portion, similar to fig. 15, more magnified. х 100. Fig. 17. Diagrammatic figure of the first cortical cells. Fig. 18. Diagrammatic figure of the commencing growth of the cortical cells from the ramification of the ** third? (r с") ; “cort. сей 1” indicates the first lateral cell, and “ cort. cell 2” one of the frontal cells, which lay the foundation of the cortex, indicated in figs. 15 and 16. Piare XXIX. Fig. 1. Tuft of Ballia scoparia, of natural size. Fig. 2. A portion of a plant, x 50. Fig. 3. Diagrammatic figure of the extremity of a branch, showing two spine-like thickened acuminate branchlets, the apical cell also spine-like, thickened aud acuminate; the pits and stoppers very minute and nearly obsolete. | Fig. 4. Mode of branching of cortical cells in this species. Fig. 5. Commencement of a branch by the outgrowth of the lateral ramification-cell (r c). Fig. 6. A stem or branch with a secondary branch, showing edge view of two ramification-cells (rc, rc). Figs. 7, 8. The mode in which the branches appear as if inserted. Fig. 9. А ramification-cell (ғ c), seen from the front, showing its elliptic figure (lenticular in edge view, see fig. 6) and central pit between it and the parent joint. Figs. 10 and 11 show the mode of giving off of the dependent * cortical” cells from the principal stems : r с, the ramification-cells ; 2, the septum cutting off the lenticular cell (marked ?), of questionable purpose. In fig. ll, at а occurs one of the ultimately numerous branches going to make up the confused plexus seen in some specimens. They frequently show here and there the formation of the lenticular ramification-cells. It would have been better if fig. 11 had been inverted to indicate the dependent direction of these cortical ramifications. Fig. 12. Ballia callitricha (typical) : nat. size. Fig. 13. A spray of this beautiful plant, х 50, showing its elegant closely set “ plumes.” Fig. 14. Showing cortical investment : x 25. Fig. 15. Same more highly magnified: x 100. Fig. 16. Portion of Ballia Brunonia, of nat. size. : Fig. 17. Part of the same specimen, x 50. (Mr. Eaton's plant seems to accord more with this form.) Fig. 18. Small portion, still more highly magnified: x 100. P" Тав.28. "> Trans: Linn. Soc. Бако Bor ot W. Archer. «я \ CCEEEXY 7 a е? w B t Tas. 29 . Trans. Linn. Soc. SER. 2.Вот. Vor I E . W. Archer. К LOR AG mic em vi. эы uL d ac БЕ Аж Po Ж ТНЕ TRANSACTIONS OF THE LINNEAN SOCIETY оғ LONDON. SECOND SERIES.—BOTANY. VOLUME I. PART THE FIFTH. LON DON: PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON-HOUSE ; AND BY LONGMANS, GREEN, READER, AND DYER, PATERNOSTER- BOW. M.DCCC.L XXVIII. ХІУ. ХУ. ХУТ. XVII. ЖҮЗІП; XIX. CON TENTS. PART V.—Janvary, 1878. Оп the Geographical Distribution of the Meliacez;. Ву M. CASIMIR DE CAN- DOLLE, of Geneva. (Communicated бу G. Вехтнам, Esq., F.R.S., V.P.L.S.) page 233 New British Lichens. By the Rev. W. А. LEIGHTON, В.А. Camb., Ғ.Г.4., FBS. Ed., $c. i : Ы 237 New Irish Lichens. By the Rev. W. A. LEIGHTON, В.А. Camb., Ғ..8., Ғ.В.8. Ed., $c. ом Report on the Liliacee, Iridacee, Hypovidacee, and Hemodoracee of Wel- witsch’s Angolan Herbarium. Ву J. G. BAKER, Esq., F.L.S.. 245 Contribution to the Lichenographia of New Zealand. Ву Онавтвв Кмент, Esq., F.L.S., Auditor-General of New Zealand 275 On some Points in the Morphology of the Primulacee. By M. T. MASTERS, MD, IPIS, ELB “Ғе. 52... 285 288 ты 7 XIV. On the Geographical Distribution of the МеНасез. Ву M. CASIMIR DE КО of Geneva. (Communicated by G. Вехтнам, Esq., F.R.S., V.P.L.S.) (Tab. & Map=Plates XXX. & XXXI.) | Read April 19th, 1877 Havine just completed a revision of the whole order of Meliacez, it has occurred to me that a general survey of their geographical distribution would not be devoid of interest. It would certainly be premature to aim at any thing like precise lines of de- marcation of the areas of the species themselves in the case of plants so thoroughly exotic, considering that, most likely, a great many new species or new stations of known specific forms remain yet to be discovered. But, on the other hand, we have good reason to suppose that very few genera will have to be added to those already admitted. When A. de Jussieu, in the year 1830, published his classical monograph of Meliaceæ and Cedrelacez, he had to deal with 135 species, which he distributed into 34 genera. At the present time our herbaria contain more than 560 species, forming only 35 genera. It is true that 10 of the generic names originated by Jussieu have been given up since, as founded on insufficient grounds, whilst more modern authors have created as many as 12 new genera. But it is to be remarked that those 12 new genera, taken together, do not include more than 38 distinct species ; consequently more than 380 new forms belong to genera already admitted by A. de Jussieu. It is, in particular, worthy of notice that South America, which has been so much explored by modern bota- nists, has yielded only one new genus of Meliacez since A. de Jussieu's time, that genus (.Elhuteria) being hitherto represented by a single species. This being the case, it seems that a study of the geographical distribution of the genera of that family is not uncalled for; and I will now proceed to state at once the principal results to which such an investigation has led me. Nearly all МеШНасеге actually known inhabit the countries situated between the tropics, beyond which their number is found to decrease very rapidly. South of the equator their area seems to be limited to the 30th degree of latitude, the only exceptions that I know of being Trichilia peruviana, which has spread as far south as Chile, and Sirtocqium spectabile, the single representative of the order in New Zealand on the 40th degree. In our hemisphere spontaneous Meliaceze have not hitherto been observed north of the tropics, with the exception, however, of Cedrela sinensis and Aglaia odorata, which have been collected in the neighbourhood of Pekin, on the 40th degree, and two species of Melia, indigenous in the southern part of Japan. Moreover it is well known that Melia Azedarach, a plant most likely derived originally from India or Tropical Africa, has from time immemorial been introduced into all countries bordering on the shores of SECOND SERIES.—BOTANY, VOL. I. 2L 284 M. CASIMIR DE CANDOLLE ON THE GEOGRAPHICAL the Mediterranean, while Cedrela sinensis is now cultivated in the open air, in the Jardin des Plantes at Paris. The existence of these few eccentric stations, in both hemispheres, clearly demonstrates that the fact of the Meliaceæ being confined to intertropical or subtropical regions is not to be accounted for solely by the present climatic conditions, and must, to a great extent, have resulted from other more remote causes of a geological nature. But this is a consideration on which I need not dwell, since it is so thoroughly in accordance with the modern notions of geographical botany. A general view of the distribution of Meliacez is shown in a map (Pl XXXI.) accompanying this paper, the coloured and shaded portions of which indieate the regions where that order is more or less represented. А dark green line, drawn round the outer extremities of those regions, embraces their whole area, including even such eccentric and exceptional stations as Pekin or New Zealand. А first inspection of this map would rather lead to considering South America the principal centre of habitation for the Meliacex, as they here occupy a larger con- tinuous area than іп any other part of the world. But it will presently be shown that the study of the geographical distribution of all the genera. composing the family points decidedly to a totally different conclusion. In order to arrive at a satisfactory conception of the case as it stands, I have drawn up a synoptic Table (Pl. XXX.) of the number of species of each genus in all the countries where Meliacez are known to exist. | On the left of the Table is a list of the genera, 35 in number, following each other according to their morphological affinities; whilst in the upper part I have entered the names of all the various countries which are to be taken into account. Those names are written in their geographical order from west to east, beginning with Chile on the left, and ending with the Navigators' archipelago on the right side of the Table. For the sake of greater precision, the prineipal provinees of South America are recorded separately one from the other, Africa and Australia being also subdivided into less extensive regions. Тһе western peninsula of India, or India proper, is, for the same reason, kept apart from Birmania and Malacea, which are grouped together with Cochin- china. Finally, more distinct, but relatively small, regions or islands have also been grouped together, such as Columbia with Venezuela, Borneo with Celebes, &с. Before taking into consideration the contents of the synoptie Table, it is necessary to offer a few remarks concerning the nature of the various genera to which it refers. It will be observed that, with the single exception of 4zadirachto, A. Juss., all of them have been admitted by Dr. Hooker, in Bentham and Hooker's ‘Genera Plantarum,’ as well as subsequently by Dr. Hiern, in his important contribution to the flora of British India. On the other hand, in conformity with Miquel’s and with Dr. Baillon's views, some . few of the generic names contained in the above-mentioned ‘Genera Plantarum ° have 3 been given up. I have, however, retained the genus Walsura, which Dr. Baillon has joined to Неупеа, and the name Dysoxylum, which corresponds to his Epicharis. Thus "ë find E it, with very few exceptions, all the genera referred to in my synoptic Table DISTRIBUTION OF THE MELIACE. 235 have, independently of my own researches, and after careful scrutiny, been also admitted by the highest authorities on the subject. Consequently I do not hesitate to consider them as being thoroughly natural. To return to the figures contained in the Table, it is evident that each of the horizontal series formed by them furnishes us with a complete statistical representation of the geo- graphical distribution of one genus ; at the same time a continuous thick, dark line drawn round the extremities of all such horizontal series enables us to understand clearly the relation which exists between that geographical distribution and the morphological affinities of the various genera. A first result becomes thus manifest. We are at once struck by the fact that the greatest number by far of the genera of Meliacez are to be found in the central part of the Table, whilst оп the sides there are only а few сепегіс forms presenting but a slight mutual affinity. In other words, we may say that the American and East-Polynesian genera, rare and very distinct, are morphologically connected by a series of numerous intermediate forms, all concentrated iu the Asiatie regions or in their neighbouring archipelagoes. 16 is, for instance, a positive fact, that out of the 35 genera composing the whole order, no less than 26 are represented within the limits of the western coast of India on one side, and of the eastern shore of Australia on the other. We also observe that, with the exception of Melia and Cedrela, these various genera are very much localized, the figures corresponding with each of them forming only short horizontal series in the synoptic Table. With respect to the genus Мейа, its exceptionally large area is more apparent than real, being due, for the most part, to the wide spread- ing of a single species, namely, Мейа Azedarach, which is not to be considered indigenous out of Tropical Africa and India. I have consequently left all the other stations of that species outside the continuous line including the genus, which is thus reduced to a very moderate area. The genus Cedrela, on the other hand, extends in reality over a vast surface ; but it must be observed, at the same time, that there exists a striking morphological difference between the American and Indian or Australian species, a fact which could not, however, be made apparent in the Table. Another general conclusion to be derived from the synoptic Table is the great analogy Which exists between the Polynesian and the Asiatic Мейасее. In fact, we see that all the Polynesian species, excepting Vavea Amicorum, belong to what may be termed the Indian type of Meliacess—that is to say, to genera which are represented in either one or the other of the Indian peninsulas. : With regard to Australia, the case is somewhat different. Besides may — be- longing to the Indian type, that continent contains also two genera ( Owenia and Lansium) which belong exclusively to it; and moreover ten out of the eleven species composing the genus Flindersia are hitherto known only in Australia. Although the whole tribe of Meliez, so abundant in Tropical Africa, is entirely missing in South America, there _ 655, nevertheless, a striking analogy between the types of those two distant con- _ tinents. Not only are the three genera Trichilia, Сагара, and even Guarea represented in Africa as well as in South America, but, moreover, Carapa procera is actually common to the west coast of Africa and the Guianas. 286 ON THE GEOGRAPHICAL DISTRIBUTION OF THE МЕШАСЕЖ. No less remarkable is the total difference existing between all the American Meliaceæ and those of the most eastern parts of Polynesia,—not a single American species, not a single American genus, being hitherto represented further west than the westerly slopes of the Peruvian or Columbian Andes, and all Asiatic genera having, as above mentioned, not been found further east than the Navigators’ Islands. t is, however, to be observed that the Brazilian genus Cabralea posseses the closest affinities with the genus Dysoxylum, which. is spread all over Polynesia—a fact which might, perhaps, suggest some amount of analogy between the American and the Indian types; but at the same time it is to be observed that the genus Cabralea is exclusively confined within the southern provinces of Brazil, none of its species having hitherto been found on the western side of the Andes. I pointed out at the beginning of this paper how premature it would be to take into serious consideration the exact boundary-lines of the species of Meliaceze. Some general remarks, however, bearing on this part of the subject are not altogether out of place. As might have been expected, the insular species are in general very much localized. A striking illustration of this consists in the fact that 18 out of 19 species of Dysoxylum inhabiting New Caledonia are special to that island, a single one of them existing also in Australia and in other parts of Polynesia. On the other hand, some species are remarkable for the great extent of their areas. For instance, Sandoricum indicum, (Maingayia) mar- ginatum, and Chisocheton paucijugum, which are all of them indigenous in the Malayan peninsula, have also been found in Java, and even as far east as Borneo. Again, Tri- chilia havanensis, Guarea filiformis, 6. trichilioides, Swietenia Mahagoni, and Cedrela Glaziovi are common to the West Indies and several provinces of South and Central America. | In conclusion, the chief features of the geographical distribution of МеНасеге may be summed up briefly as follows :— 1. The number and the mutual affinities of the various genera of Meliaceze are found to decrease gradually from the Asiatic regions towards Africa and America on one side, and towards Eastern Polynesia on the other. 2. There exists a marked analogy between the American and the African Meliacee, whilst all the Polynesian species and genera of the order belong to the Indian type. 3. New Caledonia contains a remarkable number of very distinct species, all of which, except опе ( Flindersia Fournieri), belong to the Indian type, though they are cain to be found in that island. 4. In Australia we find three Indian genera, Turrea, Dysoxylum, and Cedrela, asso- ciated with the three genera Owenia, Synowm, and Flindersia, which, with the exception of a single species (Flindersia Fournieri), belong exclusively to that country. 5. No species of Meliaceze have hitherto been collected іп the most eastern islands of Polynesia—that is to say, further east than the Navigators’ group. It is, however, hardly credible that the order should be entirely absent from the Sandwich or Galapagos archi- pelago. Should subsequent explorations reveal the existence of Meliaceze in those islands, it will be Е to know whether they belong to the Indian or to the American Trans. Linn. Soc. Ser. 2. Вот. Vor. I. Tas. 30. TABULAR ARRANGEMENT ILLUSTRATING THE GEOGRAPHICAL DISTRIBUTION OF THE GENERA AND SPECIES OF THE MELIACEJE, ACCORDING TO M. CASIMIR DE CANDOLLE. 'Spu€[ST $10j921AU NT "frr "розой ләү "зтиоретв)) MONT "wTerjsny JSE 1 '®їүвпү Чао "soume AoN чошт, "вооот puer sourddrnq "uude f "eun 'Soqopo)) рив oou1oq "BARS "елтіп "вт -шцәо/) рив ‘gome tieu. "uvurepuy "по|Хә/) “(враяптаәй пләҙвәм) отра 'snrunej раз uoqinoq “ов рив avoseSepryy ЕЕ она. БЕ £13 "вомру [eordoag, 7804 'ә@вг) 59 | 64 "woupy [TBordory, 489 №. 'Sorpug 389A, тосле "5В ИТП) ‘OTXIN вомәшү [81909/) "eurueq "еЈәпләпә д ров vrqurn]o;) "плод "em ER] 7 1 2 9 4 ZR [18] 15 3 3 1 2 6 30 6 | 10 1 үрү 1 8 8 | 11| 48 | 15 14| 2 | 4 г OIPADESBHA {se ix Поти. МОКОН vv МҰРА i-e L AZADIRACHTA . С SANDORICUM ........ САЙҒА А; DYBOXYLUM с... CHISOCHETON ...... Goi Ко т S ANON 1.13... БАНТОМ... BEDDOMEA..... SYNÓUM о... OWENIE (=. 2 DasycoLEUM........ MEENSERHGTA n VAVREA .. s stis TRIOGHEEA 2... Hari .... vV WWIIIHODBRLL "WINSBLHIMS ‘Wut -HACE | C.De.Candolle. Trans. Linn Soc. SER. 2. Вот. Vor, І Tas. ЗІ. 0 y 20_ 40 60 80 : à 100 B 120 < 140 2 160 180 160 140 20 0 Y (ШЕК: i 1 os wt- ` A. no Е : = à % > „же р Ко хал n Q. Cha ДІ ES ^acojnox| чә | | — | ы т dijo et. of Cancer Smiri Is | E t AC d [uen de "Sambia I* а ‘Tropic of Capricorn . UM Seems о + ИННИ 60 А- M À—nMM———— ———— "T ШШ? WO 154! JD Шы n Sod Ro | | 21 Cirele 4. > 4 $: ES kiana TA | к | — & | | Г ON MERCATORS PROJECTION | | | | | | Shewing 2. | | | | , ТНЕ SIX ZOOGEOGRAPHICAL REGIONS j | | ——— 6 ae r^ | | ri | 60 Longe 80 of Greenwich 100 120 140 160 180 160 мо 120 100 тыы. ga "pe LE vem LEO =m E 1 —= wim bo ЕЕЕ | 76 = 20 ; о IE - Stanford? Grog" Estal" London GEOGRAPHICAL DISTRIBUTION OF THE MELIACEJE [ 237 1 " XV. New British Lichens. By the Rev. W. А. LEIGHTON, B.A. Camè., F.L.S., F.B.S.Ed., $c. (Plate XXXII.) Read December 7th, 1876. LECIDEA. тмрквкззїтА, Leight. Albido-cinerascent, thin, filmy, rimulose, nigro- limitate. (K yellow, then red.) Apothecia black, minute, punctiform, numerous, crowded into small groups of 3 or 4 or more, impressed or sunken in the thallus, each apothecium circumscissed so as to appear surrounded by a thin thalline margin ; hypothecium thin, nigro-fuscous; paraphyses indistinct, apices colourless; hyme- nium tinged with brown; spores 8, fuscous, rotundo-oblong, 1-septate, more or less constricted in the middle. On bluish-grey slates, rare, 1876. | Gelatina hymenea I dirty blue. Spores :014—0145 mm. long, "009 mm. broad. Several plants grow on the same stone, and by the union of their dark hypothalli map out the surface as in 7. geographica. Allied to L. stellulata, Tayl, but distinct by the thin filmy thallus, larger spores (one third larger), and the singularly impressed apothecia. Llanachar bridge! and Fort Hill!, near Fishguard, Pembrokeshire (1876); Festiniog road, near Dolgelley !, North Wales. Moel-y-gest! near Tremadoc. LECIDEA rEPHRIZANS, Leight. Thallus nearly obliterated, only a few whitish depressed scattered areole remaining; (К — C — ;) hypothallus cinerascent, predominating ; apothecia black, numerous, plane or slightly concave, prominent, sessile, moderate ; margin thickish, eventually obliterated; hypotheciwm nigro-fuscous, subtended by а pale greyish-blue hyaline excipulum ; hymenium thin, pale greyish-blue ; paraphyses distinct but cohzerent, apices nigro-fuscous ; spores 8, colourless, narrowly ellipsoid, simple, minute. On hard slaty rocks, rare, 1876. Spores ‘009-01 mm. long, -004—005 mm. broad. Spermatia minute, shortly cylin- drical, straight. Gelatina hymenea I dirty blue. Llanachar bridge! near Fishguard, Pembrokeshire (Sept. 1876). LECIDEA RYSSOLEA, Leight. Albido-cinerascent, thick, tartareous, diffracto-areolate ; areole plane or subplano-convex, irregularly wrinkled, nigro-limitate. (K yellow, then red.) Apothecia nigro-fuscous, opake, somewhat rusty, large, very prominent and sessile, numerous, margin thick, paler, entire or flexuose ; hypothecium very thick, fusco-nigricant; hymenium thin, brownish; paraphyses indistinct, apices . thick, nigro-fuscous ; spores 8, JSuscous, oblong, 1-septate, small. On Caradoc sandstone rocks, rare, 1876. е SECOND SERIES.—BOTANY, VOL. 1. 2M E. 7 238 THE REV. W. А. LEIGHTON ОМ NEW BRITISH LICHENS. Spores "016-017 mm. long, ‘007-008 mm. broad. Gelatina hymenea I pale dirty blue. Medulla I—. Fort Hill! near Fishguard, Pembrokeshire (Sept. 1876). LECIDEA IMPONENS, Leight. Thallus obsolete; apothecia black, minute, numerous, scattered, sessile, plane or subconcave, margin thin, eventually obliterated ; pale bluish-cinerascent within; hymenium and hypothecium hyaline, pale bluish-cine- rascent; excipulum horny, bluish-cinerascent, externally darker ; paraphyses thick, coherent, apices nigricant ; spores 8, colourless, elliptical, simple. Parasitic on Lecanora polytropa (Ehrh.), rare, 1876. Spores "014-015 mm. long, "0055 mm. broad. Gelatina hymenea I dirty fuscescent. Fort Hill! near Fishguard, Pembrokeshire (Sept. 1876). LECIDEA GLAUCOMARIA, Nyl. Thallus obsolete, merely a small irregular, deformed, pale nigro-fuscous patch or discoloration, on which are the apothecia; apothecia fusco- nigrous, opake, clustered, subinnato-sessile, plane, margin thickish, paler, entire, often subflecuose; hymenium fuscescent; paraphyses indistinct, coherent, apices incrassate, fusco-nigricant ; hypothecium nigro-fuscous, thin; spores 8, colourless, oblongo-ovoid, 3-septate. Parasitic on Lecanora glaucoma (Hoffm.), rare, 1876. * Gelatina hymenea I pale blue, then vinous red. Spores 021—025 mm. long, "008—009 mm. broad ” (Nyl. Obs. Holm. 177, fig. 10; Scand. 245). The apothecia some- times form warty agglomerations. Spores "0235 mm. long, "009 mm. broad. Not to be confounded with Arthonia glaucomaria, Nyl. (A. varians, Dav.), which grows on the apothecia (not on the thallus) of Lecanora glaucoma (Hoffm.) This lichen has been found sparingly in France and at Stockholm, but has been now detected for the first time in Great Britain. Goodwick Bay! Pembrokeshire (Sept. 1876). Ls&CiDEA PARELLARIA, Nyl. (in Flora, 1876, p. 239). Thallus none; apothecia black, opake, small, scattered, single or 2 or 8 together, sessile chiefly but not exclusively т or near the margins of the areole of the matrix; thallus plane, margin thickish, prominent, more or less pruinose ; hypotheciwm nigro-fuscous, thin; paraphyses indistinct, conglutinate, apices clavate, nigricant; spores 8, fuscous, lineari-oblong or oblong, generally 1-septate, sometimes 2-septate, and sometimes 3-septate in the same apothecium ; gelatina hymenea I bluish, then fuscescent. Parasitic on Lecanora parella (L.), rare, 1876. Spores 011—014 mm. long, "005 mm. broad. This interesting lichen has been found once in the West of Ireland, by Mr. Larbalestier in 1876; and 16 now occurs very plenti- fully on stone walls about Fishguard, Pembrokeshire (Sept. 1876). THE REV. W. A. LEIGHTON ON NEW BRITISH LICHENS. 239 LECIDEA ENDOMEL&NA, Leight. Pale cirereo-virescent, opake, glebulose; glebule scat- tered or areolato-diffract, thick, convex, composed of minute conglomerate convez, roundish, or sublobate subfurfuraceous squamules. (K pale yellow, C pale yellow.) Apothecia violet-black, rather large, innato-sessile, at first plane with a thickish margin, afterwards very convex and immarginate, slightly pruinose, concolorous within; hypothecium very thick, nigro-fuscous, with a paler brown enclosed inter- mediate stratum ; spores 8, colourless, small, elongato-ellipsoid, simple. On stone walls, very rare, 1874. A very distinct species. Hymenium palish blue, I dense violet. Paraphyses cohzerent, apices fuscous. Spores 0115—0125 mm. long, '0045-'005 mm. broad. ODONTOTREMA MAJUS, Leight. Whitish or evanescent; apothecia brownish, prominent, : scattered, rotundate, suburceolate ; margin thickish, prominent, persistent, nearly entire; hypothecium colourless or slightly fulvescent; paraphyses distinct, slender ; spores 8, colourless, broadly lineari-oblong, 3-septate. On drift wood and decorticated stumps, rare, 1876. ; I renders the linear азс} and spores lutescent. Spores double the size of those of О. minus, Nyl., 019—021 mm. long, "0075-0085 mm. broad. Spermogonia brownish, oblong, minute; spermatia oblong or lineari-oblong, “0095 mm. long, "008 mm. broad. This I possess amongst the Arctic American lichens of Sir John Richardson ; and now (Sept. 1876) it has occurred on drift wood іп Goodwick Bay, Pembrokeshire. It has also been detected on old willow stumps at Hencote Pool, near Shrewsbury, and on wood chips in Gloucestershire. Spores of О. minus, Nyl. (see fig. 15), 0115-014 mm. long, 005mm. broad. Spores of О. longius, Nyl. (see fig. 17), ‘0195-021 mm. long, -005--007 mm. broad. VERRUCARIA NEOTTIZANS, Leight. Apothecia black, minute, clustered, hemispherico- conical; peritheciwm carbonaceous, entire; paraphyses distinct, slender ; spores 4, Suscous, lineari-oblong, 3-septate, constricted. Parasitic on the thallus of Beomyces rufus, DC., very rare, 1876. Gelatina hymenea I—. Spores “0245-025 mm. long, ‘0085-009 mm. broad. Llanachar bridge! near Fishguard, Pembrokeshire (Sept. 1876). қ VERRUCARIA FUMOSARIA, Leight. -Apothecia black, minute, hemispherico-depressed, semi-immersed ; epithecium poriform ; perithecium dimidiate, much incurved at the base ; paraphyses none; spores 8, fuscous, broadly or rotundo-oblong, 1-septate. Parasitic on the thallus of Lecidea Suscoatra, f. fumosa, Ach., rare, 1876. Spores -0155—016 mm. long, “007 mm. broad. Pen Cow! near Fishguard, Pembrokshire (Sept. 1876). E. ҮлььтсавтА PERTENUIS, Leight. Cinereo-nigricant, thin, filmy, continuous, opake, Surfuraceous ; apothecia black, opake, minute, numerous, hemispherical, sessile ; 2M 2 940 THE REV. W. А. LEIGHTON ON NEW BRITISH LICHENS. perithecium dimidiate, black ; epithecium a minute роге; paraphyses distinct, long, thickish ; spores 8, colourless, elongato-acicular, slender, 3-septate. On slaty rocks, rare, 1876. Differing altogether from V. chlorotica (Ach.) in the attenuated acicular spores and other characters, as above. Gelatina hymenea I lutescent, especially asci and spores. Spores 034-035 mm. long, 004 mm. broad. Spermogonia numerous, minute, verru- carioid; spermatia excessively minute, oblong or lineari-oblong, straight. Goodwick Вау! Pembrokeshire (Sept. 1876). DESCRIPTION OF PLATE XXXII. Fig. 1. Lecidea impressula. Vertical section of an apothecium, enlarged. Fig. 2. Spores, magnified 1200 times. Fig. 3. Vertical section of the enlarged apothecium of Lecidea tephrizans. Fig. 4. Spores of L. tephrizans, magnified 1200 times. Fig. 5. Lecidea ryssolea, apothecium in vertical section, enlarged. Fig. 6. Spores of same, magnified 1200 times. Fig. 7. The enlarged apothecium of L. imponens, in vertical section. Fig. 8. Spores of L. imponens, magnified 1200 times. Fig. 9. Lecidea glaucomaria, vertical section of enlarged apothecium. Fig. 10. Spore of same lichen, magnified 1200 times. Fig. ll. L. parellaria, enlarged apothecium in vertical section. Fig. 12. Spores of L. parellaria, magnified 1200 times. Fig. 13. Lecidea endomelena. Ароһһесішп in vertical section, enlarged. Fig. 14. Its spores, magnified 1200 times. Fig. 15. Spores of Odontotrema minus, magnified 1200 times. Fig. 16. Spores of O. majus, magnified 1200 times. Fig. 17. Spores of O. longius, also magnified 1200 times. Fig. 18. Spores of Verrucaria neottizans, magnified 1200 times. Fig. 19. Spores of V. fumosaria, magnified 1200 times. Fig. 20. Spores of V. pertenuis, magnified 1200 times. W.ALeighton del. TRANS.LiNN.Soc.Sken.2.Bor VoL 1 Тав 32. Leculea umpressuca, L ryssol Mas . L.tephrizams L.Dapearcza L. glaucomaria 5 — tà L.endometena.. Odoniotrema long tins Vfumosaru.. О minus é Ж Verrucaria pertenuxis ке 4 Г neottizaas. DBlair lth. NEW BRITISH LICHENS. M&NHanhart imp [ 941 1 XVI. New Irish Lichens. By the Rev. М.А. LEIGHTON, B.A.Camb., F.L.S., F. B.S.Ed., $c. (Plate XXXIII.) Read April 5th, 1877. PERTUSARIA INCARNATA, Leight. Thallus creamy-white, thin, tartareous, effuse, smooth, plane, minutely rimuloso-diffract. (К — О —.) Apothecia single, of a waxy appearance, paler than the thallus, prominent, sessile, depresso-globose, variable in size according to age, in maturity large; ostiolum yellowish flesh-colour, lecanoroid, at first small and round, poriform, eventually dilated and urceolate, surrounded by a broad, thick, pale, waxy, (Ла іше margin ; hymenium in section ceraceo-gelatinose, flesh-coloured; paraphyses slender, more or less branched and entangled, apices colourless; spores 2, colourless, very large, lineari-oblong or ellipsoid, simple, with a very narrow limb or epispore. On granite rocks, very rare, Kylemore, co. Galway, Ireland, Mr. Larbalestier (1875). Spores 062—067 mm. long, ‘083—085 mm. broad. Gelatina hymenea I violet. This beautiful lichen is excessively rare, and apparently quite new to science. It has occurred only once to Mr. Larbalestier, whose indefatigable and skilful lichenological researches in the west of Ireland have added so many and such rare novelties to our lichen- flora. In general external aspect it a good deal resembles Lecanora epulotica, Ach., for which it might cursorily be readily mistaken; but the number, size, and shape of the spores and other characters keep them abundantly apart. It seems allied to Pertusaria rhodoleuca, Th. Fries; but that species is at once distinguished by its 8 smaller spores. ^ Pertusaria rhodocarpa of Garovaglio’s ‘Comment. de Pertusariis Europ. Med.’ p. 8, t. iv. f. 4, is like it in the large size and narrowly limbed two spores, but differs in growing on mosses, and having a verrucoso-granulose thallus and coacervate aggregate numerous apothecia. Garovaglio’s plant, which he refers doubtfully to P. rhodocarpa, Krbr. (Syst. р. 384), cannot be identical with that German lichen, inasmuch as Th. Fries, who has examined an original specimen from Körber, ‘in Lich. Scand.’ p. 322, refers Kórber's lichen to Varicellaria rhodocarpa, Th. Fr., which has an immense single 1-septate spore, and is represented in Th. Fr. L. Scand. Exs. 73! and has reaction K—, C crimson, and is altogether different in its sorediate thallus. ШЕСІрЕА РОГОЗРОВА, Leight. Thallus white or albo-glaucous, very thin and filmy, effuse, smooth, and somewhat shining, unequal or rugulose (K— C—); apothecia black, minute, scattered, adnate, plane, with a narrow slightly prominent margin, at length plano-convex and immarginate; hypothecium nigro-fuscous; hymenium 242 THE REV. W. А. LEIGHTON ON NEW IRISH LICHENS. colourless; paraphyses distinct, thickish, apices globular, nigro-fuscous; spores 8, elliptical, fuscous, with a paler fuscous roundish cell at each apez. On old thorn-bushes, very rare. Ballinahinch churchyard, co. Galway, Ireland, Mr. Larbalestier (1877). Much resembling in general aspect Lecidea myriocarpa, DC. Gelatina hymenea І deep blue. Spores “02—022 mm. long, ‘(009 mm. broad. The peculiar shape and structure of the spores keep it as an entirely distinct species. LECIDEA GRUMOSA, Leight. "Thallus evanescent; apothecia rufo-fuscescent, minute, scat- tered, adnato-sessile, plano-convex, the slight margin soon obliterated ; hypothecium colourless; hymenium yellow, grumous; paraphyses altogether indistinct, apices colourless ; spores 8, in lineari-obovate asci, colourless, oblong, simple, filled with very minute spherical granules. On pine-bark, rare. Ballinahinch, co. Galway, Ireland, Mr. Larbalestier (1877). A very singular species. Gelatina hymenea I blue, then fulvescent. Spores :013- "015 mm. long, :007—009 mm. broad. LECIDEA ANTROPHILA, Larbal. "Thallus ochraceo-virescent, thin, effuse, pulverulento- Surfuraceous (K— С —); apothecia fulvo-rufescent, small, scattered, excessively convex and prominent, sessile, immarginate, а very pale narrow margin only appear- ing when wetted ; hypothecium darkly fulvo-rufescent, thick; hymenium colourless ; paraphyses indistinct, apices colourless ; spores 8, colourless, linear, or lineari-oblong, simple, very minute. | On the interior of caves, rare. Miderlin! near Kylemore, co. Galway, Ireland, Mr. Larbalestier (1877). ; Gelatina hymenea I very pale blue. Spores "009 mm. long, '0025 mm. broad. A very remarkable and apparently distinct species. CALICIUM KYLEMORIENSE, Larbal. Parasitical on a whitish thallus, apparently that of Lecanora cinerea, L.; apothecia excessively minute, shortly stipitate ; capitulum erateriform, nigro-fuscous, somewhat shining ; stipes very short, of a paler whitish colour; paraphyses entangled, branched; spores fuscous or fusco-nigrescent, sphe- rical, with a distinct narrow epispore, variable in size. On shady rocks, rare. Miderlin! near Kylemore, co. Galway, Ireland, Mr. Larba- lestier (1871). · Spores 006—008 mm. diam. А very beautiful new species. VERRUCARIA LARBALESTIERUL, Leight. Thallus albido-glaucescent, thin, tartareous, effuse, minutely rimuloso-diffract ; areole plane, smooth, somewhat shining, having l or 2 apothecia in the centre of each areola; (К yellow, O'— ;) apothecia black, moderately large, semi-immersed, internally dark, nigro-fuscous; ostiolum hemi- THE REV. W. А, LEIGHTON ON NEW IRISH LICHENS. 243 spherical, prominent, shining; epithecium a minute depressed pore; perithecium | black, entire; paraphyses попе; spores 8, fusco-nigricant, oblong, 1-septate, minute. On granite rocks, rare. Near Kylemore! co. Galway, Ireland, Mr. Larbalestier (1877). A delicately beautiful species, to which I give the name of its talented discoverer. Spores “016 mm. long, ‘008-0085 mm. broad. Gelatina hymenea I —. Fig. 1. Fig. 2. Fig. 3. Fig. 4. Fig. 5. Fig. 6. Fig. 7. Fig. 8. Fig. 9. Fig. 10. Fig. 11. Fig. 12. Fig. 13. Fig. 14. Fig. 15. Fig. 16. Fig. 17. DESCRIPTION OF PLATE XXXIII. Pertusaria incarnata. Enlarged view of the thallus, to show the peculiar structure, the position, and the nature of the apothecia. Vertical section of apothecium of the same, enlarged. Spores of the same lichen, magnified 600 times. Lecidea polospora. Vertical section of apothecium, enlarged. Its paraphyses, also very considerably enlarged. Spores of the same lichen, magnified 1200 times. Lecidea grumosa. Vertical section of an apothecium, enlarged. Ascus of L. grumosa, enlarged. Spores of the same lichen, magnified 1200 times. Lecidea antrophila. Section of apothecium, enlarged. Spores of L. antrophila, magnified 1200 times. Calicium kylemoriense. Three of the apothecia, of natural size. An apothecium of the same, enlarged. Spores of this lichen, magnified 1200 times. Verrucaria Larbalestierii. Enlarged view of the thallus, to show its structure and the position of the apothecium. Vertical section of the apothecium, also enlarged. Spores of the same lichen, magnified 1200 times. WALeighton del TRANs.LiNN Soc. Sen 2 Bor Vou | ТАВ 33 RR : 2 МА N. Hanhart imp: a NEW IRISH LICHENS ПРИ 5 | 245 | XVII. Report on the Liliacee, lridacee, Hypowidacee, and Нетодотасее of Welwitsch’s Angolan Herbarium. Ву J. G. Влкіж, Esq., F.L.S. (Plates XXXIV.-XXXVI.) Read March 15%, 1877. THe present paper is devoted to a systematic account of the Liliacez, Iridaceæ, Hypoxidaceze, and Hæmodoraceæ of the Angolan herbarium of the late Dr. Welwitsch. As will be seen, a very large proportion of the species are new to science. In the .Liliaeez there are two new genera, represented by five species; but the remaining uovelties all belong to generic types already known either in Central Africa or at the Cape. Тһе previously known flora of Central Africa in these orders is poor in numbers, and does not possess any strongly marked individuality of character. Its members, with few exceptions, belong either to the great cosmopolitan or great widely spread exclu- sively Old-World genera, like Ornithogalum, Scilla, Urginea, Chlorophytum, Asparagus, Dracena, and Gladiolus, or are out-wandering representatives of the well-known cha- racteristic Cape types, such ав Aloë, Kniphofia, Eriospermwm, Аиса, Morea, Ano- matheca, Geissorhiza, Aristea, and Lapeyrousia. Three well-known genera which attain their maximum in Central Africa are Dracena, Sanseviera, and Gloriosa. Wal- leria, with two species, is the only previously known endemic Central-African genus. Dr. Welwitsch’s collection, though it will more than double the number of species known in Central Africa, does not materially alter these general characters. In Liliacew it adds the two new endemic genera to which I have just referred; it carries within the tropic four Cape genera not previously known as intertropical—Sandersonia, Tulbaghia, Schizobasis, and Haworthia—and plants in more firmly Aloë and Eriospermum as inter- tropical genera. In the other orders all the genera represented are known in Central Africa already. I need scarcely add my tribute of appreciation of the excellence of Dr. Welwitsch’s specimens, and of the pains which he must have taken in selecting them to show all the various stages and characters of the plants as fully and clearly as possible, or explain that in drawing up these descriptions I have been in many cases greatly indebted to the notes which he made upon the plants at the time when they were gathered. In my ‘ Systema’ I have already indicated the position in the sequence of species of the Iridacez of this collection. In the present paper І have done this for the Liliaceze by means of the numbers prefixed to the names, which refer to my mono- graphs already published in the Journal of the Linnean Society. My best thanks are due to the executors of the late Dr. Welwitsch for the facilities they have given me in consulting the specimens and notes, which latter have been separated and copied for the British Museum, and upon both of which this paper is founded. SECOND SERIES.— BOTANY, VOL. 1. 2N 246 MR. J. G. BAKER ON THE LILIACE# ETC. Order LILIACES. Кхтрвоғтл, Mench. (Journ. Linn. Soc. xi. 360.) К. BENGUELLENSIS, Welw. Herb. Herba 3-5-pedalis, rhizomate verticali lignoso, fibris copiosis cylindricis. Folia 3—4-pedalia, linearia, basi 6-9 lin. lata, a basi ad apicem attenuata, acuminata. Scapus 2-3-pedalis. Racemus subspicatus densus semi- pedalis vel pedalis, expansus 8-9 lin. crassus. Pedicelli brevissimi vel subnulli. Вгасбеге deltoidee membranacez, perianthio triplo breviores. Perianthium luteum campanulatum 25 lin. longum, segmentis oblongis tubo zequilongis. Stamina demum exserta, antheris oblongis i lin. longis. Capsula globosa brunnea rugosa 21 lin. longa, seminibus brunneis lucidis. Ни” а in palustribus juxta rivulos, socialis cum Richardia angolensi её Epilobio æthiopico. Floret Dec.-Jan. К. ANDONGENSIS, Baker. Herba gigantea, rhizomate incrassato, fibris radicalibus сат-. nosis pedalibus vel sesquipedalibus. Folia linearia 6—9-pedalia, basi 8-9 lin. lata, ad apicem sensim angustata. Scapus foliis zquilongus, basi 1 poll. crassus. Racemus densus, inexpansus semipedalis. Pedicelli brevissimi. Вгасбеге lanceo- latee 2 lin. longs. Perianthium rubellum oblongum 3 lin. longum, segmentis oblongis tubo turbinato equilongis. Flores perfecte expansos non vidi. Pungo Andongo in dumetis petrosis alte graminosis inter Luxillo et Cotete. Floret Maio. * Pongo" incolarum. Species angolenses ab reliquis notis recedunt perianthio pro- funde lobato. TULBAGHIA, Linn. (Journ. Linn. Soc. xi. 369.) 1%, T. xqQuiNocTIALIS, Welw. Herb. Bulbus ovoideus 6-9 lin. crassus, tunicis brunneis 10%. membranaceis. Folia 3-5 anguste linearia glabra subglaucescentia erecta 4-6 poll. longa, vix 1 lin. lata. Scapus gracilis folia superans. Umbella 3-flora, spathe valvis 2 lanceolatis 3-4 lin. longis, pedicellis horizontalibus vel cernuis 3-4 lin. longis. Perianthium album 5 lin. longum, segmentis lanceolatis tubo :equilongis. Corona minutissima flava profunde lobata. Pungo Andongo, in dumetis hwmidius- culis. Floret Martio. Dircapr, Medic. (Journ. Linn. Soc. xi. 395.) D. охутовом, Welw. Herb. Bulbus subglobosus 6-9 lin. crassus, tunicis mem- branaceis albidis. Folia 2 suberecta subulata glabra facie canaliculata 4-6 poll. longa $ lin. lata. Scapus gracilis flexuosus 6-9-pollicaris. Racemus laxus secundus 3-3-pollicaris 5-10-forus. Pedicelli 1-2 lin. longi. Bracteze lanceolato-deltoidex . euspidatze viridulze 2-3 lin. longe. Perianthium viridulum cinnamomeo-rubens, seg- ~ mentis interioribus 5 lin. longis, exterioribus 7-8 lin longis apice caudatis. Pung? as eee declivis breviter herbidis rupium ipsius Presidii, regio temp. (2400-3500 м * D Аы Baker i in Jonh. Linn. Soc. xi. 400. Бени secundus laxe 6-15- 17*. 28". 144, ОҒ WELWITSCH'S ANGOLAN HERBARIUM. 247 florus. Capsula globoso-quadrangulata 4 lin. longa. Huilla, regio temperata, in dumetis herbidis prope Lopollo. Floret Nov.—Dee. i D. nATERITIUM, Welw. Herb. Bulbus globosus 9-12 lin. crassus, tunicis mem- branaceis albidis. Folia 2-4 linearia carnoso-herbacea erecta glabra ad apicem attenuata, pedalia vel sesquipedalia, 5—6 lin. lata. Scapus fragilis gracilis pedalis vel sesquipedalis. Racemus laxus secundus 2-6-pollicaris. Pedicelli ascendentes, inferiores demum 6-9 lin. longi. Bractew lineares acuminate 3-6 lin. longs. Perianthium lateritium, segmentis interioribus 4—44 lin. longis, exterioribus semi- polliearibus apice distincte caudatis. Capsula sessilis 11 lin. longa. Huilla, regio temperata, in rupestribus editis Marro de Monimo ditionis Lopollo. Floret Martio. Dircapi comosum, Welw. Herb. (Plate XXXIV. figs. 1-3.) Bulbus depresso- globosus, tunicis læte incarnatis, basi bulbillis stipatus. Folia linearia carnoso- herbacea glabra erecta bipedalia, basi 5-6 lin. lata, ad apicem attenuata, facie canaliculata. Scapus 3-4-pedalis, raro 5-pedalis. Racemus 11-2-pedalis laxe 30- 60-florus, pedicellis inferioribus 3-4 lin. longis, floribus viridibus, inferioribus per- fectis 8-9 lin. longis, caudis brevibus, centralibus segmentis interioribus 3—4 lin. longis, exterioribus longe caudatis, floribus superioribus more Muscari parvis abortivis. Bracteæ superiores lineares acuminate 3—4 lin. longæ. Ни? а in arenosis | dumetosis siccioribus inter Lopollo et Catumba. Floret Aprili. UnGiINEA, Steinh. (Journ. Linn. Soc. xiii. 215.) U. (SQUILLA) PsrLosTACHYA, Welw. Herb. Bulbus magnitudine pugni, apice hypogzeo, tunicis multis albidis coriaceis. Folia post scapum perfecta 5-6 sub- coriacea lorata glabra pedalia vel sesquipedalia 12-14 lin. lata. Scapus 2-3-pedalis. Racemus laxus sesquipedalis, expansus 1 poll. latus, pedicellis suberectis vel cernuis 8-4 lin. longis. Bracteæ minutæ lineares supra basin calcarate. Perianthium albidum 21 lin. longum, segmentis late viridi vittatis. Stamina perianthio squi- longa, filamentis subulatis, antheris oblongis flavis 2 На. longis. Stylus cylindricus exsertus. Capsula globoso-trigona 4 lin. longa. Cazengo т dumetis siccis arenosis. Floret Aug. . U. (Атвисорзїз) comosa, Welw. Herb. Bulbus magnus globosus, tunicis mem- branaceis. Folia 6-8 carnoso-herbacea lineari-lorata erecta glauca 2—3-pedalia, basi 9-12 lin. lata, facie canaliculata, ad apicem attenuata. Scapus tripedalis et ultra. Racemus pedalis vel sesquipedalis, expansus 2-23 poll. crassus, rachi stricta in- crassata rugosa. Bractes subulate 6-9 lin. longæ ante anthesin comosze. ; Pedicelli inferiores 6-9 lin. longi. Perianthium albidum 4-5 lin. longum, segmentis lanceo- latis viridi-vittatis 5-nervatis. Filamenta perianthio paulo breviora, antheris flavis oblongis 1 lin. longis. Capsula 9 lin. longa profunde 3-lobata, seminibus multis discoideis nigris nitidis 4 lin. latis. J/ossamedes et Huilla, in pascuis sylvaticis humidiusculis inter Lopollo et Monino. Floret Dec.-Jan. 2х2 248 30*. 37*. 46*. MR. J. С. BAKER ON THE ІЛІЛАСЕ Ж ЕТС. URGINEA (ALBUCOPSIS) CHLORANTHA, Welw. Herb. Bulbus globosus 12 poll. crassus tunicis membranaceis. Folia 6-8 carnoso-herbacea lorata glabra bipedalia basi 1 poll. lata ad apicem acutum attenuata. Scapus elongatus. Racemus florifer semipedalis, fructifer sesquipedalis 11 poll. latus. Bracteæ subulate 6-9 lin. longs ante anthesin comosz, haud calcaratæ. Pedicelli 4-6 lin. longi. Perian- thium albidum 44 lin. longum, segmentis lanceolatis viridi vittatis dorso distincte 5-nervatis. Filamenta linearia perianthio paulo breviora, antheris oblongis 1 lin. longis. Stylus ovario zquilongus. Capsula depresso-globosa 9 lin. lata profunde lobata, seminibus in loculo multis discoideis nigris nitidis 4 lin. latis. Loanda, Pungo Andongo et Golungo Alto, in declivis olim cultis. Floret Jan.—April. OnNITHOGALUM, Linn. (Journ. Linn. Soc. xiii. 257.) O. (BERYLLIS) BENGUELLENSE, Baker. Bulbus ovoideus 6-9 lin. crassus, tunicis membranaceis brunneis apice haud setiferis. Folia 2—8 subteretia carnoso-herbacea glabra erecta semipedalia 1 lin. crassa. Scapus gracilis 6-9-роШсатів. Racemus laxus lanceolatus subsecundus 3—4-pollicaris. Pedicelli ascendentes, inferiores 3-4 lin. longi. Bractez minute deltoidex, inferiores calearat:e. ^ Perianthium 2 lin. longum, segmentis albis purpureo vittatis. Stamina perianthio paulo breviora, antheris conformibus linearibus. Capsula immatura oblonga. Huilla, regio tem- perata, in dumetis arenosis ad Lopollo. Floret Nov. О. (BERYLLIS) CEPZFOLIUM, Baker. Bulbus globosus tunicatus 11-2 poll. crassus. Folia 4—5 lanceolata glabra acuminata pedalia vel sesquipedalia, basi 1-12 poll. lata, facie canaliculata. Scapus bipedalis. Racemus densus 2-3-pollicaris, bracteis inferioribus multis haud floriferis. Bractee lineares 5-6 lin. longs haud cal- -caratee, ante anthesin comose. Pedicelli 2-3 lin. longi. Perianthium campanu- latum З lin. longum, segmentis albis dorso viridi vittatis multinervatis. Filamenta conformia lanceolata 2 lin. longa, antheris oblongis 1 lin. longis. Capsula globosa ‚ 4lin.longa, seminibus in loculo 6-8 compressis dimidiato-oblongis nigris. Pungo 18°. Andongo et Huilla, in apricis herbidis prope Lopollo. Floret Dec. SciLLA, Linn. (Journ. Linn. Soc. xiii. 228.) S. (EuscrLLA) HISPIDULA, Baker. Bulbus oblongus 2-3 poll. crassus basi durus, tunieis densis crassis coriaceis. Folia 5-6 lanceolata rigida dura coriacea glabra 4-6 poll. longa, basi 1 poll. lata, nervis crebris exsculptis. Scapus 4-6-pollicaris cylindricus sursum pilosus. Racemus densus, expansus 4-8 poll. longus, 1 poll. erassus, rachi incrassata rugosa dense pilosa. Pedicelli patentes 3—4 lin. longi. Втасбеге lineari-subulatze, pedicellis zequilongs. Perianthium campanulatum al- bidum 14 lin. longum. Stamina inclusa conformia. Stylus exsertus. Capsula _ sessilis, globosa, profunde lobata, 2 lin. longa, seminibus in loculo binis erectis.— Huilla, in dumetis apricis arenoso-argillaceis prope Lopollo. Floret Nov. 40*. 40%, 40%. 47%. OF WELWITSCH'S ANGOLAN HERBARIUM. 249 SCILLA (LEDEBOURIA) BENGUELLENSIS, Baker. Bulbus globosus 1-2 poll. crassus, tunicis brunneis. Folia 2—6 sessilia synanthia linearia suberecta 4 lin. longa medio 8—4 lin. lata ad basin et apicem attenuata. Scapus flexuosus 2-3-pollicaris. Racemus densus oblongus 12-18 lin. longus, floribus inferioribus cernuis, pedicellis 2-3 lin. longis, bracteis minutis deltoideis. Perianthium campanulatum 14 lin. longum, segmentis rubro-purpureis late viridi vittatis. Capsula sessilis globosa. иа in pascuis apricis subarenosis prope Lopollo, Floret Nov. S. (LEDEBOURIA) PoLYPHYLLA, Baker. Bulbus globosus, 1 poll. crassus, tunicis brunneis. Folia 10-12 synanthia erecta subteretia glabra 3 poll. longa vix 1 lin. crassa. Scapus foliis sequilongus. Racemus laxus pauciflorus 5-6 lin. longus. Pedicelli perianthio sequilongi. Bracteze minutz deltoideæ. Perianthium cam- panulatum intense rubro-purpureum 15 lin. longum, segmentis oblanceolatis infra medium reflexis. Stamina perianthio equilonga. Pungo Andongo et Golungo Alto in declivis rupestribus. Floret Julio. Habitus Scille autumnalis. S. (LEDEBOURIA) ARENARIA, Baker. Bulbus ovoideus;6-9 lin. crassus, tunicis brun- neis membranaceis. Folia 3-6 synanthia linearia erecta carnoso-herbacea glauco- viridia obseure petiolata 2 poll. longa 12 lin. lata ad apicem sensim angustata. Seapus tripollicaris erectus gracilis. Racemus laxus oblongus, ехрапвав 9-12 lin. longus, 5-6 lin. latus, pedicellis 11-2 lin. longis, inferioribus cernuis, bracteis minutis deltoideis. Perianthium campanulatum purpureum 1 lin. longum, segmentis infra medium faleatis. Filamenta perianthio zequilonga saturate purpurea. Cap- sula globosa stipitata. Pungo Andongo in pascuis arenosis sylvaticis. Floret Martio. S. (LEDEBOURIA) SIMIARUM, Baker. Bulbus ovoideus 9-12 lin. crassus, tunicis brunneis membranaceis. Folia 3—4 synanthia carnoso-herbacea lanceolata glauco- viridia patula basi purpureo maculata 3-4 poll. longa, supra basin 5-6 lin. lata, acuta, haud petiolata. Scapus 2-3-pollicaris. Racemus oblongus sublaxus, florifer expansus 12-15 lin. longus, pedicellis 13-2 lin. longis plerisque cernuis, bracteis minutis deltoideis. Perianthium viridulum purpureo tinctum campanulatum 1 lin. longum. Filamenta cyanea perianthio demum zequilonga. Capsula sessilis globosa. Pungo Andongo et Loanda in sabulosis estate inundatis. Floret Aprili. “Bulbi simiis gratissimum pabulum." 47*. S. (LepEBoURIA) FLACCIDULA, Baker. Bulbus globosus 8-9 lin. crassus, tunicis brunneis membranaceis. Folia 2-4 synanthia flaccida lorata carnoso-herbacea glabra 6-12 poll. longa 4-8 lin. lata acuta sessilia. Scapus debilis foliis equilongus. Racemus laxus oblongus, fructifer 3—4 poll. longus, 6-8 lin. latus, pedicellis cen- tralibus 2-3 lin. longis, bracteis minutis deltoideis. Perianthium campanulatum viridulum 11—11 lin. longum,—segmentis infra medium falcatis. Stamina perian- 250 MR. J. С. BAKER ON THE LILIACE# ETC. thio sequilonga. Stylus exsertus. Capsula sessilis globosa. Pungo Andongo in rupestribus ipsius Presidii. Floret Nov. | 47%. SCILLA (LEDEBOURIA) LAXIFLORA, Baker. Bulbus globosus 12-15 lin. crassus, tunicis brunneis membranaceis. Folia 3-4 synanthia carnoso-herbacea oblanceolata 4-5 poll. longa, supra medium 5-6 lin. lata, acuta, ad basin angustata. Scapus foliis eequilongus. Racemus cylindricus laxissimus 4—5-polliearis, pedicellis centralibus patentibus 3—4 lin. longis, inferioribus cernuis, bracteis minutis deltoideis. Peri- thium viridulum campanulatum 1 lin. longum,—segmentis infra medium faleatis. Stamina perianthio z:equilonga. Ovarium sessile, stylo exserto. Loanda in collinis breviter graminosis. Floret Maio. 48*. S. (LEDEBOURIA) coNGESTA, Baker. Bulbus globosus 12-2 poll. crassus, tunicis brunneis membranaceis. Folia 3 synanthia carnoso-herbacea lanceolata sessilia acuta 3—6 poll. longa, prope basin 12-15 lin. lata. Scapus 3-4-polliearis. Racemus densus subspicatus, florifer 2 poll. longus, pedicellis infimis 1 lin. longis, bracteis minutis deltoideis. Perianthium campanulatum viridulum 21 lin. longum. Geni- talia inclusa. Ovarium sessile. Аиа in collinis dumetosis agri Lopollensis. Floret Jan. 48*. S. (LEDEBOURIA) PLATYPHYLLA, Baker. Bulbus magnus ovoideo-globosus. Folia 4 synanthia carnoso-herbacea glabra oblonga acuta semipedalia, medio 15-18 lin., basi 1 poll. lata. Scapus semipedalis. Racemus densus 11-2 poll. longus, pedi- cellis omnibus ascendentibus 12-2 lin. longis, bracteis minutis deltoideis. Perian- thium viridulum campanulatum 2 lin. longum. Ми а, regio temperata, in dume- tosis collinis parce graminosis inter Lopollo et Catumba. Floret Dec. 57. S. (LEDEBOURIA) LANCEÆFOLIA, Baker in Saund. Ref. Bot. t. 182. Species capensis late disseminata. Lopollo in collinis et Ниша in pascuis arenosis sylvaticis prope Mompulla, regio temperata. Floret Oct. ALBUCA, Linn. (Journ. Linn. Soc. xiii. 285.) 14. *. A (FALCONERA) MYOGALOIDES, Welw. Herb. Bulbus ovoideus 1 poll. crassus, tunicis d apice in setas dissolutis. Folia 3—4 synanthia glabra. carnoso-herbacea erecta a anguste linearia 1—2-pedalia, 3—4 lin. lata. Scapus pedalis, vel sequipedalis. Race- mus laxus 6-12-florus semipedalis vel pedalis, pedicellis ascendentibus inferioribus . 2-8 poll longis, floribus erectis, bracteis lanceolatis inferioribus 6-9 lin. longis. ce Ium album 6-7 lin. longum, segmentis 2 lin. latis dorso viridi vittatis. = E Filamenta exteriora antheris parvis instructa. Stylus prismaticus ovario æqui- longus. Capsula oblonga perianthio æquilonga. Huilla in rupestribus editis Morro | : de Monino, Floret Tan Ad А. caudatam Em men arcte accedit. 10*. 14*. 14% ОҒ WELWITSCH'S ANGOLAN HERBARIUM. 251 ALBUCA (FALCONERA) MONOPHYLLA, Baker. A. Juncifolia, Welw. Herb., non Baker in Gard. Chron. 1876, 534. Bulbus globosus 6—9 lin. crassus, tunicis teneris membra- naceis apice haud setiferis. Folia solitaria raro bina teretia carnoso-herbacea glabra pedalia vix 1 lin. erassa glauco-viridia basi canaliculata. Scapus 4-—-8-pollicaris. Racemus laxe 2-6-florus, pedicellis erectis 3-4 lin. longis, floribus subnutantibus, bracteis lanceolatis cuspidatis 3—4 lin. longis. Perianthium 41-5 lin. longum luteum, segmentis late viridi vittatis, exterioribus 11 lin. latis. Filamenta omnia fertilia, antheris parvis oblongis luteis. Stylus prismaticus ovario sequilongus. Pungo Andongo in rupestribus editis Presidii Pedra de Cabonda. Floret Aprili. A. (PALLASTEMA) SUBSPICATA, Baker. Bulbus globosus 6-9 lin. crassus, tunicis membranaceis apice haud setiferis. Folia 3-4 synanthia subulata carnoso-her- bacea pedalia ad apicem attenuata. Scapus flexuosus 4-6-роШсатів. Racemus angustus 2-4-роШсатів, 4—12-florus, pedicellis adscendentibus 11-3 lin. longis, floribus erectis vel patulis, bracteis lanceolatis longe cuspidatis inferioribus 6-9 lin. longis. Perianthium album 6 lin. longum, segmentis viridi vittatis 2 lin. latis. Filamenta omnia fertilia perianthio paulo breviora. Stylus gracilis ovario equilongus. Hwuilla ad rupes editas de Morro de Mino. А. (PALLASTEMA) GALEATA, Welw. Herb. Bulbus ovoideus, tunicis brunneis mem- branaceis. Folia synanthia glabra carnoso-herbacea flaccida viridia bipedalia 3—4 lin. lata, facie canaliculata. Scapus 1-2-pedalis. Racemus laxus 3—5-pollicaris, expansus 13-2 poll. latus, pedicellis erecto-patentibus, infimis 5-6 lin. longis, floribus superioribus patulis, inferioribus cernuis, bracteis lanceolatis cuspidatis 3-6 lin, longis. Perianthium albidum subbilabiatum 7-74 lin. longum, segmentis late viridi vittatis, tribus inferioribus declinatis, suprema galeata. Stamina omnia fer- tilia, antheris oblongis 11 lin. longis. Stylus filiformis ovarium superans. In dumetis petrosis prope Giucanda. Floret Sept. “ Dispositio laciniarum perigonii singularis quasi corollam bilabiatam fingens. Lacinia interna suprema galeam mentiens, in coneavitate stamina 2 lacinia arcte adpressa gerentem." Welw. Herb. 15. A. ANGOLENSIS, Welw Baker in Saund. Ref. Bot. t. 336. Huilla т pratis altis PE 1e. herbidis inter Lopollo et rivulo ejusdem nominis sporadica. Floret Dec.—Feb. А. (PaLLAsTEMA) CHLORANTHA, Welw. Herb. Bulbus globosus 13 poll. crassus, tunicis membranaceis apice haud setiferis. Flora plura synanthia lorata carnoso- herbacea glabra acuminata 14-2-pedalia, basi 1% poll. lata. Scapus elongatus. Racemus subdensus pedalis, expansus 2 poll. latus, pedicellis 13—2 lin. lon gis, bracteis lanceolatis acuminatis 12-2 poll. longis, floribus inferioribus horizontalibus, supremis ascendentibus subsessilibus. Perianthium flavum 9-12 lin. longum, segmentis dorso late viridibus 2—8 lin. latis. Filamenta omnia fertilia, antheris oblongis 2 lin. 252 MR. J. С. BAKER ON THE LILIACE/E ETC. longis. Stylus filiformis 5-6 lin. longus, ovario zquilongus. Pungo Andongo in dumetis ad ripas fluminis Juanza. Floret Martio. Draca#na, Vand. (Journ. Linn. Soc. xiv. 523.) 11%. D. PARVIFLORA, Baker. Arbor 20-35-pedalis dichotome ramosa, ramis erecto- patulis. Folia dense rosulata sessilia lanceolata subcoriacea 3-4-pedalia, medio 3 poll., supra basin 1 poll. lata, costa preter apicem perspicua, marginibus concolo- ribus. Panicula pedalis et ultra, floribus densis, ad apices ramorum 20-30 con- gestis, bracteis perminutis deltoideis, pedicellis 2 lin. longis. Perianthium albidum 4—45 lin. longum, segmentis tubo zquilongis. Stylus exsertus. Golungo Alto ad rivulos in editioribus sylvaticis. Floret Feb. “ Die calunga" incolarum. 15. D. FRAGRANS, Gawl. in Bot. Mag. t. 1081. Pungo Andongo. 19*. D. nrrens, Welw. Herb. Arbuscula 7-10-pedalis. Folia linearia laxe disposita subpatula subcoriacea viridia lucida 6-10 poll. longa, 41-6 lin. lata, sessilia acuta, costa preter apicem perspicua, marginibus concoloribus. Pedunculus brevissimus. Panieula semipedalis vel pedalis, ramis pluribus brevibus patulis, pedicellis brevis- simis, bracteis minutis deltoideis. Perianthium cylindricum albidum rubro tinctum, segmentis ligulatis tubo brevioribus. Stylus exsertus. Golungo Alto et Pungo Andongo in montosis rupestribus. Floret Sept. 25*. D. AcAULIs, Baker. caulis, rhizomate lignoso 4—1 poll. crasso. Folia omnia in rosulam basalem congesta lanceolata acuta firma 3-6-pedalia medio 2 poll. lata, in petiolum pedalem alatum sensim angustata. Panicula pedunculata 4-5-pedalis, ramis multis flexuosis ascendentibus racemosis, bracteis inferioribus semipedalibus. Pedicelli brevissimi aggregati. Perianthium albidum 14-15 lin. longum, segmentis tubo paullo longioribus. Stamina segmentis æquilonga, antheris oblongis 13 lin. longis. Stylus demum exsertus. Golungo Alto, regio 2 (1000—2400 pedes) т sylva- ticis editioribus et ad rivulorum latera montium de Alta Queta. Floret Julio. 33*. D. моховтаснул, Baker. Acaulis. Folia omnia in rosulam basalem congesta oblongo-lanceolata acuta semipedalia vel pedalia, infra medium 13-2 poll. lata, ad basin euneatam rotundata, petiolo pedali instrueta. Pedunculus pedalis, bracteis multis parvis lanceolatis preeditus. Racemus simplex densus erectus 2—3-pollicaris, pedicellis brevissimis fasciculatis, bracteis minutis deltoideis. Perianthium non vidi. Васса profunde lobata. Тазша Principis in herbidis subhumidis sylvarum non procul oceano. Sept. (in fructu); et planta affinis Angolensis in ditione Golungo _ Alto adest foliis magis oblongis 6-8 poll. longis, 2-23 poll. latis basi magis rotun- . datis, pedunculo breviore et racemo laxiore. 38*. D. INTERRUPTA, Baker. Arbuscula subscandens 3—4-pedalis, ramis virgatis eylin- ricis gracilibus. Folia in rosulas segregatas disposita oblongo-oblanceolata cusp ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 253 data 4-5 poll. longa, medio 2-23 poll. lata, basi cuneata in petiolum brevem 12 lin. latum angustata, costa preter apicem perspicua. Racemus simplex breviter pedunculatus 4—5-pollicaris, floribus dense fasciculatis, pedicellis 12 lin. longis supra medium articulatis, bracteis membranaceis deltoideis vel lanceolatis 2-3 lin. longis. Flores odorati albi. Perianthium cylindricum 9-11 lin. longum, segmentis tubo subduplo brevioribus. Filamenta segmentis paullo breviora, antheris 1 lin. longis. Stylus demum exsertus. Pungo Andongo, in sylvis ad marginem fluminis Tombe. Floret Martio. SANSEVIERA, Thunb. (Journ. Linn. Soc. xiv. 546.) 3*. S. BRACTEATA, Baker. Folia rigidissima coriacea lanceolata albo maculata rubro 28*. 38*. marginata 1-2-pedalia, medio 2-21 poll. lata, basi in petiolum latum elongatum angustata. Pedunculus crassus strictus sesquipedalis vel pedalis, foliis multis reductis lanceolatis, inferioribus 8-4 poll. longis, preditus. Racemus simplex densus semipedalis, bracteis lanceolatis membranaceis 9-12 lin. longis, pedicellis brevissimis fasciculatis. Perianthium album 21—3 poll. longum, segmentis ligulatis 8-9 lin. longis. Stamina inclusa, antheris oblongis 11 lin. longis. Pungo Andongo, in rupestribus subsiccis editioribus. Floret Jan. Ad S. longifloram arcte accedens. ' ASPARAGUS, Linn. (Journ. Linn. Soc. xiv. 594.) А. EQUISETOIDES, Welw. Herb. Frutex ramosissimus, ramis erecto-patentibus pinnatis, ramulis gracilibus erecto-patentibus pedalibus vel semipedalibus, internodiis ultimis 2-3 lin. longis. Folia ramorum basi caleari parvo deltoideo instructa. Cladodia ad nodos swpissime nulla, raro producta solitaria subulata 2-3 lin. longa ad ramos adpressa. Flores non vidi. Pungo Andongo, in arenosis ad ripas fluminis Cuensa. : A. BENGUELLENSIS, Baker. Arbuscula decumbens glabra subscandens, ramis graci- libus copiose pinnatis, ramulis adscendentibus vel patulis 4-6 poll. longis, inter- nodiis ultimis 2-3 lin. longis. Folia inferiora lanceolata, obscure calcarata. Cladodia 3-6na subulata ascendentia 6-9 lin. longa. Pedicelli 1-2ni, raro 3-4ni, 4-5 lin. longi, medio articulati. Perianthium infundibulare З lin. longum, segmentis ob- longo-spathulatis diu ascendentibus brunneo vittatis. Antheræ minute oblongz, filamento triplo breviores. Mossamedes et Huilla, in dumetis sylvestribus ad Mum- pullo. Floret Julio. . А. PSILURUS, Welw. Herb. Herba erecta, 2-3-pedalis, habitu 4. officinalis, vamis et ramulis copiosis gracillimis erecto-patentibus, ultimis pedalibus vel semipedalibus. Folia ramorum calcari obscuro deltoideo deflexo instructa. Cladodia 6-8na, subu- . lata erecto-patenta 4-9 lin. longa. Pedicelli cernui solitarii 2-3 lin. longi medio _ SECOND SERIES.—BOTANY, VOL. I. 20 254 2” MR. J. G. BAKER ON THE LILIACEÆ ETC. articulati ad nodos efoliatos sepe producti. Perianthium infundibulare 13-2 lin. longum, segmentis albis brunneo vittatis diu ascendentibus. Antherz minute ob- longe, filamento triplo breviores. Pungo Andongo, in dumetis Cuazensibus, frequens. Floret Maio. . ASPARAGUS DEFLEXUS, Baker. Frutex ramosissimus, glaber, ramis gracilibus, inferio- ribus deflexis, ramulis multis brevibus szepe internodio unico solum preditis, inter- nodiis ultimis 1-12 lin. longis. Folia ramorum calcari parvo deflexo pungente in- structa. Cladodia subulata 6-10па 8-4 lin. longa. Pedicelli 13-2 lin. longi, medio articulati, ad nodos laterales 1-2ni, ad nodos terminales ssepe 4—6ni. Perianthium 15 lin. longum, segmentis oblongo-spathulatis brunneo vittatis. Stamina perianthio distincte breviora, antheris minutis oblongis. Pungo Andongo. Floret Oct. . А. PUBESCENS, Baker. Frutex ramosissimus pubescens, ramulis arcuato-ascenden- tibus semipedalibus vel pedalibus, internodiis ultimis 12-9 lin. longis. Folia ramo- rum basi calcari deltoideo parvo pungente instructa. Cladodia 4-8na, subulata, erecto-patentia glabra 2-3 lin. longa. Pedicelli 1-2ni erecto-patentes 14-2 lin. longi, infra medium artieulati, ad ramos vel nodos cladodiis nullis preeditos ssepis- sime producti. Perianthium 14 lin. longum. Anthere minute oblonge lutem. Васса magnitudine pisi parvi. Huwilla, regio temperata, in pascuis siccis ad Catumba. Floret Dec. . А. ANGOLENSIS, Baker. Frutex ramosissimus subscandens, habitu 4. africant, ramis lignosis flexuosis puberulis, ramulis copiosis erecto-patentibus. Folia ramo- rum calcari subulato pungente deflexo 3-4 lin. longo instructa. Cladodia densis- sime fasciculata (20—50na) gracillima rigida subulata 6-9 lin. longa. Pedicelli patuli fasciculati 2 lin. longi prope basin articulati. Perianthium infundibulare 1$ lin. longum. Stamina perianthio szequilonga, antheris minutis globosis. Васса magnitudine pisi parvi. uilla, regio temperata. 73. A. AFRICANUS, Lam. Pungo Andongo, in sylvaticis. 85. A. RACEMOSUS, Willd. Loanda et Huilla, in dumetis. 90*. А. DREPANOPHYLLUS, Welw. Herb. Frutex glaber, latissime volubilis, ramis gra- cilibus lignosis, internodiis ultimis 14-2 lin. longis. Folia ramorum basi calcari pungente subulato 2 lin. longo preedita. Cladodia 3, inzequalia, linearia, falcata, rigi- dula acuta distincte costata, majora 9-12 lin., minora 2-3 lin. longa. Racemi ses- siles, 3-6 poll. longi, 44-5 lin. lati, pedicellis 3-6nis 1-1} lin. longis, medio arti- . eulatis. Perianthium campanulatum 1 lin. longum, segmentis late viridi vittatis. _ Anthere minute оой. 1 inclusee. Golungo Alto, in wee densis sporadicus. | Floret Maio. OF WELWITSCH'S ANGOLAN HERBARIUM. 255 ScHIZOBASISs, Baker. (Journ. Linn. Soc. xvi. 260.) S. ANGOLENSIS, Baker. Adenotheca aphylla, Welw. Herb. Bulbus globosus 1-14 poll. crassus intus albidus viscosus, tunicis brunneis membranaceis. Folia hysteranthia non vidi. Scapus gracilis semipedalis. Panicula deltoidea pedalis, ramis gracilibus erecto-patentibus, inferioribus compositis, ramulis ultimis 4—5 poll. longis. Pedi- celli remoti solitarii ascendentes 3—4 lin. longi prope apicem articulati. Bracteæ minutæ deltoideæ. Perianthium 2 lin. longum, segmentis oblongis brunneo vittatis. Stamina inclusa, filamentis leviter applanatis, antheris oblongis filamento brevi- oribus. Pungo Andongo, ad rupes ipsius Presidii, alt. 2400-3800 pedum. Floret Aug. ACROSPIRA, Welw. Herb. (Genus novum Asphodelearum.) Perianthium album, diu infundibulare, basi brevissime urceolatum, segmentis æqualibus oblanceolatis, nervis 8—5 in carinam centralem concretis. Stamina 6 inclusa profunde perigyna, filamentis brevibus zequalibus lanceolatis, antheris magnis linearibus basifixis apice spiraliter revolutis. Ovarium parvum sessile ovoideum 3-loculare, ovulis in loculo circiter 20 horizontalibus biseriatis; stylus leviter exsertus filiformis leviter declinatus superne sensim robustior, stigmate parvo capitato peni- сШаю. Fructum non vidi. À. ASPHODELOIDES, Welw. Herb. (Plate XXXIV. figs. 4-7.) Rhizoma obliquum, crassum, fibris multis gracilibus przeditum. Folia 5-6 sessilia subcoriacea lorata 2—3-pedalia deorsum 12—15 lin. lata ad apicem angustata glabra crebre distincte nervata. Pe- dunculus 4—6-pedalis, foliis pluribus reductis preeditus. Racemus pedalis vel sesqui- pedalis simplex vel basi ramis 1-2 parvis instructus, pedicellis ascendentibus 3—4nis inzequalibus 1-3 lin. longis, bracteis deltoideis. Perianthium 9-10 lin. longum, segmentis 1 lin. latis. Antherz lutez 5-6 lin. longe. Stylus 8-9 lin. longus. Pungo Andongo, in rupestribus editioribus, frequens. Floret Mart.—April. Dasystacuys, Baker. (Genus novum Asphodelearum.) Perianthium album diu campanulatum, segmentis æqualibus lanceolatis dorso 1-nervatis. Stamina inclusa «qualia, filamentis profunde perigynis elongatis leviter applanatis, antheris parvis oblongis versatilibus. Ovarium sessile globosum apice et lateribus profunde lobatis triloculare, ovulis in loculo paucis superpositis; stylus filiformis declinatus exsertus, stigmate capitato. Capsula globosa profunde lobata loculicide trivalvis, seminibus in loculo paucis discoideis. Herbe africane tropicales, habitu Antherici, rhizomate obliquo tuberoso, fibris multis gracilibus predito, foliis basalibus rosulatis linearibus vel lanceolatis arundinaceis crebre nervatis, pedunculo foliis reductis vaginato, floribus parvis in racemos subspicalos sepissime simplices dispositis. 1. D. ргктозтАСНҮА, Baker. Anthericum? pleiostachywm, Welw. Herb. Rhizoma tuberosum 6-9 lin. crassum, fibris pluribus radicalibus preeditum. Folia basalia 3-4 lanceolata subcoriacea glabra crebre nervata subpedalia 6-9 lin. lata. Scapus _ 2-8-pedalis, foliis multis reductis linearibus adpressis preditus. Racemi 1-3 densi uo | 202 256 MR. J. G. BAKER ON THE LILIACEZ ETC. subspieati 6-9 poll. longi 5-6 lin. lati, pedicellis brevissimis vel subnullis, bracteis setaceis 5-6 lin. longis basi dilatatis. Perianthium album 3 lin. longum. Stamina perianthio szequilonga, antheris minutis oblongis. Capsula globosa 23-3 lin. longa, lateribus profunde lobatis. Pungo Andongo, in pratis graminosis ай ripas fluminis Cuenza. Floret Aprili. 2. DASYSTACHYS FALCATA, Baker. Anthericum? faleatum, Welw. Herb. Rhizoma ob- longum tuberosum horizontale, fibris gracilibus multis preeditum. Folia basalia rosu- lata 6-8 lorata facie canaliculata basin caulis longe vaginantia superne falcata 6-9 poll. longa 10-12 lin. lata crebre distincte nervata margine crispata. Pedunculus rubellus sesquipedalis vel bipedalis, foliis multis reductis preeditus. Racemus sim- plex subspicatus, floriferus 8—4 polL, fructiferus 6-8 poll. longus, 8-9 lin. latus, pedicellis brevissimis. Perianthium album 3 lin. longum. Stamina perianthio æquilonga. Capsula 4 lin. longa, seminibus in loculo 3—4 discoideis. Huilla, regio temperata, in rupestribus Morro de .Lopollo et Mossamedes, in sylvis inter Bumbo et Bruco. Floret Oct.-Jan. 3. D, COLUBRINA, Baker. (Plate XXXV. figs. 1-6.) | Anéhericum? colubrinwm, Welw. ` Herb. Rhizoma obliquum tuberosum, fibris radicalibus multis gracilibus preditum. Folia basalia exteriora rudimentaria rubro-purpureo maculata, interiora producta 5-6 linearia glabra subcoriacea pedalia, medio 3 lin. lata, ad apicem et basin angustata. Scapus simplex pedalis, foliis 2-3 reductis przditus. Recemus sub- spicatus simplex oblongus 11-2 poll. longus, 8-9 lin. latus, pedicellis subnullis, bracteis purpureis deltoideis longe cuspidatis, infimis 3—4 lin. longis. Perianthium З lin. longum, segmentis rubro-purpureo vittatis. Stamina perianthio squilonga. Stylus declinatus longe exsertus. Capsula globosa3 lin. longa, lateribus profunde lobatis. Ни а, in pascuis rupestribus herbidis de Етраістса, alt. 5500 pedum. Floret Jan. 4. D. CAMPANULATA, Baker. (Plate XXXV. figs. 7-10.) Campylandra dasystachys, Welw. Herb. Rhizoma obliquum tuberosum, fibris multis radicalibus gracilibus proditum. Folia rosulata basalia producta 5-6 lanceolata basin caulis longe vagi- nantia pedalia vel sesquipedalia basi 1 poll. lata crebre distincte nervata ad apicem sensim angustata, marginibus minute ciliatis et crispatis. Pedunculus 3—4-pedalis, foliis multis reductis vaginatus. Racemus simplex densus subspicatus 3-6-pollicaris 8-9 lin. crassus, bracteis lanceolatis longe acuminatis, inferioribus 5-6 lin. longis. Perianthium 4 lin. longum. Stamina perianthio szequilonga. Stylus $ Ш. longus longe exsertus. Capsula 4 lin. longa, lateribus acute angulatis. uilla, regio temperata, in pratis herbidis inter Lopollo et lacum Mantalla. Floret Martio. ANTHERICUM, Linn. (Journ. Linn. Soc. xv. 290.) элө. e A. (Рилглхетом) TENELLUM, Welw. Herb. Dense ceespitosum, rhizomate apice dense setifero, fibris radicalibus cylindricis 1-2 poll. longis. Folia dense rosulata 27% ЖҮ 97°: ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 257 anguste linearia erecta glabra 2-3 poll. longa, 1 lin. lata ad apicem angustata. Scapi multi gracillimi folia заре superantes. Racemi laxissimi 2-3 poll. longi sim- plices vel paniculati, bracteis minutis deltoideis, pedicellis 14-3 lin. longis, inferiori- bus 2-3nis. Perianthium 3 lin. longum album membranaceum, segmentis anguste oblanceolatis, dorso laxe 3—4-nervatis. Stamina perianthio distinete breviora, an- theris minutis oblongis, filamentis glabris. Stylus 12 lin. longus inclusus ovario equilongus. -Ни (а, regio temperata, in pascuis humidiusculis subarenosis de Morro Monino. Floret Dec. ANTHERICUM (PHALANGIUM) DISSITIFLORUM, Baker. Rhizoma parvum tuberosum, fibris radicalibus multis gracilibus præditum. Folia linearia glabra modice firma 3—5 poll. longa, 15 lin. lata, facie canaliculata, nervis crebris distinctis12-15. Scapus supra folia eminens acute anceps. Racemus simplex laxus 1-13-pollicaris 3-6-florus, pedi- cellis solitariis 14-3 lin. longis, medio articulatis, bracteis lanceolatis acuminatis 2—8 lin. longis. Perianthium 3—4 lin. longum, segmentis oblanceolatis dorso purpureo vittatis 5-nervatis. Stamina perianthio distincte breviora, antheris oblongis 1 lin. longis, filamentis linearibus glabris. Stylus 2 lin. longus. Capsula depresso-globosa 2 lin. longa. Pungo Andongo, in pascuis humidis prope Condo. Floret Martio. A. (PHALANGIUM) BENGUELLENSE, Baker. Cæspitosum, rhizomate setis densis coronato, fibris radicalibus gracilibus. Folia plura linearia glabra canaliculata 3—6 poll. longa 3—4 lin. lata 12-15-nervata minute ciliata. Scapus anceps nudus 3-9-pollicaris, erectus vel curvatus. Рашеша deltoidea 3—4-pollicaris, floribus in- ferioribus 2-3nis, pedicellis 1-14 lin. longis, bracteis deltoideis. Perianthium 4-44 lin. longum, segmentis oblanceolatis dorso purpureo vittatis 3-5-nervatis. Stamina perianthio triente breviora, filamentis glabris, antheris parvis oblongis. Ним, regio temperata, in sylvis claris et pascuis dumosis circa Lopollo et Humpata. Floret Jan. A. (PHALANGIUM) LIMOSUM, Baker. Rhizoma tuberosum, fibris radicalibus gracili- bus, brevibus. Folia radicalia 5-6 linearia subfirma pedalia 5-6 lin. lata crebre nervata. Scapus 3-6-pollicaris. Рапісша 3—4-pollicaris, ramis pluribus erecto- patentibus, pedicellis 14-2 lin. longis medio articulatis, inferioribus fasciculatis. Perianthium 4—44 lin. longum, segmentis lanceolatis purpureo vittatis dorso distincte 3-5-nervatis. Anthere 2 lin. longs, filamentis glabris. Stylus declinatus 2-21 lin. longus. Barra Фо Bengo, in inundatis limosis eusiccatis. p / p. (PHALANGIUM) ANDONGENSE, Baker. Rhizoma obliquum tuberosum, fibris radicalibus filipendulis 2-3 poll. longis. Folia radicalia 4-6 anguste linearia glabra subcoriacea rigidula semipedalia 1-1} lin. lata crebre nervata, marginibus minute ciliatis. Scapus simplex pedalis complanatus nudus. donar simplex laxus bipollicaris, pedicellis 3—4 lin. longis infra medium articulatis, inferioribus 2-35. Perianthium 4 lin. longum, segmentis purpureo vittatis dorso 3-4-nervatis. Stamina 258 21*. di. M". 93*, MR. J. С. BAKER ON THE LILIACE/E ETC. perianthio subduplo breviora, antheris 14 lin. longis, filamento glabro duplo lon- eioribus. Capsula globosa 3 lin. longa, valvis profunde horizontaliter plicatis. Pungo Andongo, in arenosis humidis sylvaticis inter Luaillo et Cazella. Floret Jan. ANTHERICUM (PHALANGIUM) oRCHIDEUM, Welw. Herb. Rhizoma obliquum tubero- sum, setis multis coronatum, fibris carnosis filipendulis. Folia radicalia 6-8 oblongo- lanceolata subcoriacea glabra crebre nervata 6-9 poll. longa, 12-15 lin. lata. Scapus 3-8-polliearis. Racemus sæpissime simplex densus 3—4-pollicaris, pedicellis 3-6 lin. longis 3-6nis medio articulatis. Perianthium 4-45 lin. longum, segmentis ob- lanceolatis dorso purpureo brunneo vittatis 7-9-nervatis. Stamina perianthio sub- duplo breviora, antheris oblongis filamento glabro zquilongis. Stylus declinatus 3 lin. longus. Capsula globosa 3 lin. longa, valvis horizontaliter plieatis. Huilla, regio temperata, in graminosis humidiusculis ad basin Serra de Monino. Floret Jan. А. (PHALANGIUM) USTULATUM, Welw. Herb. Rhizoma gracile late reptans. Folia 3 exteriora rudimentaria lanceolata firma nigro fasciata, 3 interiora producta stricta erecta teretia glabra 6-8 poll. longa, $ lin. crassa. Scapus simplex semipedalis. Racemus laxus simplex 1-1j-pollicaris, pedicellis 2—3nis 14-3 lin. longis, bracteis purpureis lanceolatis cuspidatis, infimis 3-4 lin. longis. Perianthium 4 lin. longum, segmentis purpureo-brunneo vittatis 4—5-nervatis. Stamina perianthio subduplo breviora, antheris oblongis filamento glabro zequilongis. Huilla et Pungo -Andongo, in spongiosis sylvaticis. Floret Martio-Aprilem. А. (PHALANGIUM) CALYPTROCARPUM, Baker. Fibrz radicales cylindrice flexuose, collo radicis dense setoso. Folia producta 1-2 teretia glabra pedalia et ultra vix 1 lin. crassa. Scapus semipedalis glaber gracilis. Panicula deltoidea 6—9-pollicaris, ramis pluribus ascendentibus, racemis laxissimis, internodiis infimis 12—18 lin. longis, pedicellis 2—8 lin. longis, inferioribus geminis, bracteis minutis deltoideis cuspidatis. Perianthium 3 lin. longum tenerum, basi circumscissum, segmentis dorso fusco vittatis. Stamina perianthio triente breviora, antheris oblongis filamento glabro szequilongis. Capsula turbinata, seminibus in loculo pluribus atris minutis triquetris. Huilla, regio temperata, in pascuis petrosis tempore pluvii inundatis, inter Mumpulla et Lopollo. Floret Dec. A. (PHALANGIUM) PTEROCAULON, Welw. Herb. Rhizoma tuberosum, fibris radi- calibus gracilibus, apice setis paucis preditum. Folia 3-4 linearia subcoriacea basin caulis longe amplectentia 2-3-pedalia, 5-6 lin. lata, nervis crebris pilosis. capus anceps compressus 3-5-pedalis. Racemus simplex strictus 3—6-pollicaris. nodis inferioribus distantibus; pedicellis 3-4 lin. longis medio articulatis, inferioribus 2-3nis. Perianthium album 5-6 lin. longum, segmentis dorso anguste vittatis. t Stamina perianthio subduplo breviora, antheris oblongis filamentis linearibus glabris zquilongis. Capsula subglobosa 3—4 lin. longa. Pungo Andongo, in dumetis inter Calundo et Pedras de forage. Floret Martio. ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 259 33*. ANTHERICUM (PHALANGIUM) ARENARIUM, Baker. Fibre radicales cæspitosæ graciles semipedales filipendulz, collo radicis dense setoso. Folia producta 3-4 lanceolata faleata rigida coriacea 5-6 poll. longa, e medio 3-9 lin. lato ad apicem et basin an- gustata, пегу1 30—40 crebris dense pilosis. Scapus pedalis vel semipedalis subcom- pressus superne pilosus. Racemus densus simplex 1-2-pollicaris, bracteis pilosis lance- olatis 6-9 lin. longis, pedicellis 2-3nis medio articulatis, inferioribus 3-4 lin. longis. Perianthium album 4—4} lin. longum, segmentis dorso multinervatis. Stamina perianthio subduplo breviora, antheris linearibus filamento glabro eequilongis. Cap- sula globosa З lin. longa. Pungo Andongo, in dumetis arenosis inter Candumbe et Lombe. Floret Martio. А. pRIMIOPSIS, Baker (J. L. S. ху. 301). Planta zambesiaca a Kirkio lecta etiam ad genus pertinet. 50. А. (TRACHYANDRA) JACQUINIANUM, Schult. Huilla, regio temperata, іп pascuis et arenosis humidiusculis. Floret Jan.-Feb. Var. AFFINIS, Baker. Pungo Andongo. 54*. А. (TRACHYANDRA) PYRENICARPUM, Welw. Herb. Fibre radicales graciles, collo radicis squamis parvis membranaceis cincto. Folia 6-8 anguste linearia parce pilosa linearia facie canaliculata 6-8 poll. longa 1 lin. lata. Scapi 2-4ni 2—3-polli- cares subtiliter pilosi. Racemus laxus simplex 2—3-polliearis, pedicellis 3—4 lin. longis, inferioribus cernuis, bracteis lineari-subulatis 3—4 lin. longis. Perianthium album 4 lin. longum, segmentis oblanceolatis. Stamina perianthio triente breviora, antheris minutis. Capsula glabra turbinata 3 lin. longa, seminibus in loculo solitariis. Ниша, regio temperata, in pascuis humidis inter Humpata et Gambas. Floret Jan. 69%. А. (DILANTHES) MOLLE, Baker. Fibræ radicales dense, interdum apice filipendul oblong. Folia rosulata 20-80 petiolata; petiolus 1—3-polliearis pilis lanosis patulis densis preeditus; lamina lanceolata 2—3-pollicaris dense pilosa crebre nervata. Seapus semipedalis gracilis erectus. Racemus simplex vel furcatus densus 1-13- pollicaris, bracteis pilosis lanceolatis acuminatis 3—6 lin. longis, pedicellis 2-3 lin. longis medio articulatis, inferioribus 2-3nis. Perianthium З lin. longum, seg- mentis albis lanceolatis dorso purpureo vittatis 5-nervatis. Anthere 1} lin. longze, filamenta subulata muricata superantes. Stylus declinatus 2 lin. longus. Capsula globosa 2 lin. longa, valvis horizontaliter plicatis. Pungo Andongo, in apricis petrosis tolius Presidii. Floret Nov.-Dec. _ CHLOROPHYTUM, Ker. (Journ. Linn. Soc. ху. 321.) 5. C. MACROPHYLLUM, Aschers. С. leucolepis, Welw. Herb. Pungo Andongo, in rupium fissuris ad scaturigines et juxta rivulos Presidii. Floret Dec. 260 MR. J. G. BAKER ON THE LILIACE/E ETC. 7%. CHLOROPHYTUM FILIPENDULUM, Baker. Rhizoma obliquum, fibris radicalibus filipen- dulis 2—4 poll. longis. Folia pauca petiolata; petiolus canaliculatus 1—4-pollicaris ; lamina oblongo-lanceolata membranacea glabra 6-9 poll. longa, 11-2 poll. lata, supra viridis, infra pallida, acuta, basi cuneata, nervis perspicuis 20-25 preedita. Scapus semi- pedalis vel pedalis, nudus vel folio unico instructus. Racemus laxus semipedalis simplex vel raro furcatus, bracteis viridibus lanceolatis 4-6 lin. longis, pedicellis 2-93nis, inferioribus 8—4 lin. longis. ^ Perianthium albo-viridulum 3-4 lin. longum, segmentis dorso distincte trinervatis. Stamina perianthio triente breviora, antheris 1 lin. longis, filamento subulato. Capsula oblonga 4 lin. longa, seminibus in loculo 6-8. Golungo alto in umbrosis ad rivulos in montes de Queta. Floret Dec. 8*, C. LANCIFOLIUM, Welw. Herb. Fibre radicales graciles 3—4 poll. longæ. Folia rosulata 22*. 26*. 5-6; petiolus 5-6-pollicaris; lamina lanceolata vel oblongo-lanceolata glabra mem- branacea 3-4 poll. longa acuta, basi cuneata vel rotundata, venis perspicuis 12-15. Seapus 3-4-pollicaris. Racemus simplex laxissimus, floriferus expansus semipedalis, fructiferus pedalis, bracteis lanceolatis 8-4 lin. longis, pedicellis 3—4nis 1-14 lin. longis supra medium articulatis. Perianthium albo-viridulum 2 lin. longum, seg- mentis linearibus. Capsula globosa 2 lin. longa, seminibus in loculo tribus. Pungo Andongo, in umbrosis ad cataractas rupium ipsius Presidii. Floret Nov. C. DEBILE, Baker. Fibree radicales 2 poll. longze graciles vel incrassatee. Folia radi- calia plura petiolata; petiolus 1—3-pollicaris ; lamina lanceolata membranacea glauco- viridis glabra 2—4 poll. longa 3—4 lin. lata e medio ad basin et apicem angustata, venis perspicuis 10-12 preedita. Scapi brevissimi 2-3ni. Racemus simplex laxissimus 2-6-polliearis sepe deflexus, internodiis inferioribus 9-19 lin. longis, bracteis lan- ceolatis viridibus cuspidatis 8—4 lin. longis, pedicellis 1-2 lin. longis medio articulatis solitariis vel geminis. Perianthium albo-viridulum 2 lin. longum, segmentis line- aribus. Capsula depresso-globosa 3 lin. longa. Congo; Ambriz, ad rupes. Floret Nov. C. ANDONGENSE, Baker. Fibre radicales cylindricæ 9-4 poll. longæ. Folia radicalia 5-6 sessilia lorata suberecta glabra pedalia vel sesquipedalia medio 6-15 lin. lata chartacea, venis circiter 20, marginibus albidis. Scapus pedalis et ultra. Panicula deltoidea 1—2-pedalis, ramis erecto-patentibus, bracteis deltoideis cuspidatis 2-3 lin. longis, racemis laxis, pedicellis 3—6nis 6-9 lin. longis, medio articulatis. Perian- thium albido-viridulum 6 lin. longum, segmentis oblanceolatis dorso 5-6-nervatis. Stamina perianthio paulo breviora, filàmentis subulatis. Capsula dura globosa _ 4-5 lin. longa, seminibus in loculo 3-4. Pungo Andongo, in dumetis rupestribus omnibus Presidii. Floret Dec. C. srENOPETALUM, Baker. C. bracteosum, Welw. Herb. Huilla, т sylvis Protearum ad Lopollo. Floret Dec. ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 261 Вотвтхе, Linn. (Journ. Linn. Soc. ху. 341.) B. AsPHODELOIDES, Schult. fil. Pungo Andongo et Huilla, regio temperata, in sylvis clarioribus subarenosis. Floret Dec. = Liliacea, No. 14, Welw. Apont. p. 592. ERIOSPERMUM, Jacq. (Journ. Linn. Soc. xv. 262.) 4%, Е. srENoPHYLLUM, Welw. Herb. Tuber depresso-globosum sat magnum setis densis coronatum. Folia solitaria synanthia petiolata; petiolus 1-2-pollicaris ; lamina linearis glabra subcoriacea erecta crebre nervata 6-8 poll. longa 21-3 lin. lata, a medio ad apicem et basin sensim attenuata. Scapus gracilis semipedalis. Racemus laxus 5-6-pollicaris, bracteis minutis deltoideis, pedicellis erecto-patentibus, in- ferioribus 8-32 poll. longis, superioribus sensim brevioribus. Perianthium 2 lin. longum infundibulare albidum, segmentis ligulatis purpureo-brunneo vittatis. Sta- mina perianthio triente breviora, filamentis subulatis, antheris minutis subglobosis. Ovarium globosum sessile. Pungo Andongo, ad ripas fluminis Cumanza, in arenosis dumetosis. Floret Martio. 4%, E. ANDONGENSE, Welw. Herb. Tuber globosum 1-13 poll. crassum, setis coronatum. 12*. 15*, SECOND SERIES.—BOTANY, VOL. I. Folia solitaria hysteranthia petiolata; petiolus 2-3-pJilicaris; lamina subcoriacea oblanceolata glabra 6-8 poll. longa 1-13 poll. lata, a medio ad petiolum angustata. Scapi 1-2 palmares vel pedales. Racemi laxi 2-4-pollicares, pedicellis elongatis. Perianthium sulphureum, segmentis purpureo vittatis. Stamina inclusa, 3 interiora breviora. Ovarium sessile ovoideo-trigonum. Pungo Andongo, in dumetis frequens. Floret Nov. | Е. FLEXUOSUM, Welw. Herb. Tuber globosum 6-12 lin. crassum. Folia hyste- ranthia non vidi. Scapus flexuosus 6-9-polliearis. Racemus laxus 6-9-pollicaris 20-30-florus, гасі flexuosa, pedicellis erecto-patentibus, inferioribus 2-24 poll. longis, superioribus sensim brevioribus, bracteis minutis deltoideis. Perianthium campanulatum 3 lin. longum, segmentis ligulatis albis purpureo vittatis. Filamenta linearia perianthio triente breviora. Huilla, regio temperata, in pascuis humidiusculis inter Lopollo et Humpata. Floret Nov. E. rALuDosuM, Baker. Tuber oblongum vel globosum 9-12 lin. crassum, intus sanguineo-purpureum, fibris densis intertextis persistentibus coronatum. Folia 1-3 synanthia breviter petiolata coriacea glauco-viridia glabra lanceolata vel oblonga 1-2 poll. longa, 6-9 lin. lata, basi rotundata vel subcuneata. Beapus gracilis 4-6- pollicaris. Racemus laxus 3—4-pollicaris, bracteis minutis deltoideis, pedicellis capillaribus ascendentibus 6-12 lin. longis. Perianthium campanulatum 2 lin. longum, segmentis albidis: rubro-purpureo vittatis. Filamenta subulata perianthio triente breviora, antheris oblongis minutis. Capsula oblonga perianthio zequilonga. Pungo Andongo, frequens т paludosis. Floret Jan.—Mart. : P 262 MR. J. С. BAKER ON THE LILIACE# ETC. : 15%, ERIOSPERMUM OPHIOGLOSSOIDES, Welw. Herb. Tuber globosum, apice fibris copiosis intertextis coronatum. Folia 1—2 synanthia lanceolata glabra coriacea glauco-viridia lanceolata breviter petiolata 1-3 poll. lata, 6-7 lin. longa, basi euneata. Scapus semipedalis. Racemus laxus 3—4 poll. longus, bracteis minutis deltoideis, pedicellis erecto-patentibus, inferioribus 6-9 lin.longis. Perianthium 3 lin. longum sul- phureum, segmentis fusco-purpureo vittatis. Filamenta subulata perianthio triente breviora, antheris minutis oblongis flavis. Stylus filiformis 1 lin. longus. — Huilla, regio temperata, in pascuis humidiusculis. Floret Nov. ALLIUM, Linn. А. (RHIZIRIDIUM) ANGOLENSE, Baker. Bulbi cespitosi anguste ovoidei 3 poll. longi, 6-9 lin. lati, tunicis brunneis membranaceis. Folia 6-8 erecta anguste linearia glabra glauco-viridia carnoso-herbacea 6-9 poll. longa 14 lin. lata. Scapus validus pedalis et ultra, medio leviter incrassatus. Spathæ valv:e 2 deltoidæ 3—4 lin. longe. Umbella densa globosa 2 poll. lata, pedicellis 6-9 lin. longis. Perianthium album campanulatum 2 lin. longum, segmentis oblongis fusco vittatis. Stamina peri- anthio sequilonga, filamentis conformibus subulatis. Golungo Alto, т herbidis. Floret Maio. т GLORIOSA, Linn. С. SUPERBA, Linn. Angola. С. VIRESCENS, Lindl. Angola e£ Sierra Leone. SANDERSONIA, Hook. S. LITTONIOIDES, Welw. Herb. Tuber carnosum Gloriosa. Caulis sesquipedalis erectus, dimidio inferiore nudo flexuoso, superiore recto crebre foliato. Folia oblonga acuta subcoriacea viridia amplexicaulia subtiliter nervata 2—4 poll. longa, 1 poll. lata, apice haud cirrosa. Flores ex axillis foliorum superiorum producti solitarii, pedi- cellis cernuis 1-2 poll. longis. Perianthium aurantiaco-purpureum 9 lin. longum, segmentis lanceolatis maculis atropurpureis tessellatis, basi nectario saccatis. Sta- mina perianthio duplo breviora, filamentis filiformibus profunde perigynis, antheris versatilibus 1% lin. longis. Ovarium oblongum 3-4 lin. longum, apice obtuse lobatum; stylus erectus ovario equilongus, apice tricuspidatus. Pungo Andongo, im sylvis, alt. 2400-3800 pedum. Habitus Epipactis latifolia. WALLERIA, Kirk. W. AwGoLENSIS, Baker. Tuber carnosum difforme. Caulis rectus pedalis vel sesqui- _ pedalis, prope basin nudus. Folia 20-30 lanceolata erecto-patentia sessilia modice _ firma acuminata 3—4 poll. longa 3-6 lin. lata. Pedicelli ex axillis foliorum mul- _ torum m produ oducti, szepissime simplices 12-18 lin. longi, raro fureati, ad furcam bractea lanceolata instructi. Perianthium 6-9 lin. longum intense cyaneum, seg- ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 263 mentis lanceolatis acutis per totam latitudinem laxe nervatis. Stamina perianthio paulo breviora, antheris linearibus 4 lin. longis. Ovarium basi inferum, loculis 4-7- ovulatis. Capsula cernua. Ми а, in collinis petrosis, alt. 3800-5500 pedum. Intermediate between JF. nutans and W. Mackenzii of Zambesi-land. They may likely prove all three, as Dr. Welwitsch suggested to me, varieties of a single species. HawonTHIA, Duval. H. ANGOLENSIS, Baker. Bulbus squamis fuscis carnosis ovato-lanceolatis 15-18 lin. longis dense imbricatis preeditus, fibris radicalibus czespitosis сагпоѕіѕ Folia 6-12 synanthia rosulata erecta linearia 2-3 poll. longa glauca planiuscula, margine serrulata, dentibus minutis retrorsis. Scapus semipedalis. Racemus subspicatus laxus 4-5-pollicaris, floribus erecto-patentibus, bracteis deltoideis cuspidatis 3 lin. longis. Perianthium 6 lin. longum, segmentis ligulatis obtusis late viridi vit- tatis tubo multo brevioribus. Аа, regio subtemperata, in dumetis arenosis. Floret Nov. Атой, Linn. А. ANGOLENSIS, Baker. Subacaulis. Folia dense rosulata ensiformia bipedalia glauca immaculata 15-24 lin. lata acuminata, dentibus deltoideis patulis corneis 1 lin. longis. Seapus bipedalis, valde compressus, basi 1 poll. crassus. Racemi 1-3 densi 8-4 poll. longi, pedicellis 14-8 lin. longis, bracteis deltoideo-lanceolatis, floribus inferioribus deflexis. Perianthium sulphureum 9-10 lin. longum. Stamina in- clusa. Burro do Bengo, in collinis sylvaticis. Floret Julio. А. ANDONGENsIS, Baker. Caulis 1-2-pedalis 2- vel 3-partitus, ramis ascendentibus. Folia dense rosulata lanceolata glauca immaculata arcuato-recurvata 8-9 poll. longa, basi 15-18 lin. lata, dentibus parvis deltoideis cuspidatis corneis armata. Рапісша deltoidea, racemis densis 2-3 poll. longis, pedicellis 3-4 lin. longis, bracteis lan- ceolatis pedicello zequilongis. Perianthium 7-8 lin. longum luteo-rubrum. Sta- mina inclusa. Pungo Andongo, in rupestribus frequens. Floret Jan.—April. А. LITTORALIS, Baker. Arborescens 6—10-pedalis, trunco sepissime simplici, brachii, rarius femoris humani crassitie. Folia dense rosulata ensiformia 2-3-pedalia acu- minata immaculata 2-91 poll. lata, dentibus patulis corneis lanceolatis vel deltoideis 1 lin. longis armata. Scapus 4—5-pedalis, sursum panieulatus. Racemi pedales modice densi, pedicellis 9-4 lin. longis, bracteis lanceolato-deltoideis 4-6 lin. longis, inferioribus reflexis. Perianthium 1 poll. longum. Stamina demum leviter exserta. Loanda, totius regionis littoralis in collinis aridis frequens. Floret Maio-Julium. А. PALMIFORMIs, Baker. Fruticescens, caule 3—5-pedali, basi simplici, interdum superne ramoso. Folia dense rosulata ensiformia glauco-viridia immaculata pedalia et ultra, basi 21-24 lin. lata, dentibus deltoideis corneis patulis crebris 2 lin. longis armata. Scapus simplex vel ramosus, racemis modice densis 50-60-floris 6-8 poll. longis, 2р 2 264 МВ. J. G. BAKER ON THE LILIACE/E ETC. floribus inferioribus cernuis, pedicellis 3-6 lin. longis, bracteis minutis deltoideis cuspidatis. Perianthium splendide rubrum 9-10 lin. longum. Stamina inclusa. Huilla, regio subtemperata, т petrosis sylvaticis editioribus, Morro do Lopollo. Floret Aprili. Cum tribus preecedentibus ad stirpem A. abyssinice referendum. Атой ZEBRINA, Baker. Breviter caulescens, caule simplici semipedali 1-2 poll. crasso. Folia dense rosulata patula lanceolata acuminata glauca albo-lineata et maculis albis inordinate seriatis utrinque decorata semipedalia vel pedalia, basi 11-2 poll. lata, dentibus deltoideis corneis 12-2 lin. longis armata. Scapus 3—4-pedalis, sursum panieulatus, racemis laxis 6—12 poll. longis, pedicellis 8-6 lin. longis, bracteis lan- eeolatis acuminatis 3—4 lin. longis. . Perianthium lateritium 9-12 lin. longum. Sta- mina inclusa. Barro do Bengo, Loanda et Lebongo. Ad А. pictam capensem affinis. A. PLATYPHYLLA, Baker. Breviter caulescens, caule simplici semipedali vel pedali. Folia dense rosulata ovato-lanceolata acuminata glauco-viridia dense lineata et albo- maculata semipedalia, basi 22-3 poll. lata, dentibus deltoideis crebris corneis 14 lin. longis armata. Scapus 3—4-pedalis, racemis laxis semipedalibus vel pedalibus, bracteis lanceolatis, pedicellis 5-6 lin. longis. Perianthium rubrum 12-15 lin. longum. Stamina inclusa. Pungo Andongo, im dumetis siccis frequens. Floret Aprili. An sit varietas latifolia grandiflora prioris ? Order $ МІІАСЕ Ж. SMILAX, Linn. S. Krausstana, Hochst. Huilla, Ambaca, Benguella et Golungo Alto. Floret Feb. Order Н ®MADORACES. XEROPHYTA, Commers. X. caPILLARIS, Baker. (Plate XXXVI. fig.1). Vellosia (Ophiothamna) capillaris, Welw. Herb. Arbuscula, ramulis lignosis 5-6 lin. erassis, basibus foliorum delapsorum brunneis truncatis crebre nervatis. Folia ad rosulam producta 4-8 linearia longe aristata coriacea glabra 4-6 poll. longa 3 lin. lata crebre nervata. Pedicelli cæspi- tosi 2-3 poll. longi dense atro-viscosi. Ovarium turbinatum З lin. longum dense atro-viscosum. Perianthium albidum 15-18 lin. longum, segmentis lanceolatis acuminatis basi 3 lin. latis. Antherze lineares 5-6 lin. long, filamento brevissimo. Capsula oblonga 5-6 lin. lata dense viscosa. Ми а, in dumetis editioribus ad sylvas yode Pets et in alta pns de Empalanca, alt. 3800-5500 pedum. ж SQUARROSA, Baker. Vellosia (Ophiothamna) squarrosa, Welw. Herb. Frutex 3-4 rarius ee trunco basi 14-2 poll. crasso, basibus foliorum delapsorum sape ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 265 squarrosis. Folia producta linearia acuminata coriacea 5-6 poll. longa 5 lin. lata, marginibus superne denticulatis. Pedunculi uniflori 4—5 poll. longi superne dense viscosi. Ovarium clavatum 4-41 lin. longum densissime atro-viscosum. | Perianthii limbus 9-12 lin. longus, segmentis lanceolatis albo-ceruleis. Antheræ lineares subsessiles, perianthio duplo breviora. Capsula oblonga crustacea densissime atro- viscosa $ poll crassa. Pungo Andongo, in declivis rupestribus editioribus ipsius Presidi gregaria. XEROPHYTA VELUTINA, Baker. Vellosia (Ophiothamna) velutina, Welw. Herb. Frutex erectus 14-24-pedalis, ramis 6-9 lin. crassus, basibus foliorum delapsorum rigidis adpressis truncatis mucronatis. Folia producta ad rosulam 8-10 linearia vel lanceo- lata 6-10 poll. longa 3—6 lin. lata coriacea crebre nervata distincte costata utrinque subtiliter pilosa. Pedunculi 12-2 poll. longi, sursum atro-viscosi. Ovarium clava- tum З lin. longum dense atro-viscosum. . Perianthii limbus pollicaris cæruleus, seg- mentis lanceolatis acuminatis. Stamina perianthio subduplo breviora, antheris linearibus, filamentis brevissimis. Pungo Andongo, alt. 2400-3800 pedum, in rupes- tribus arenoso-schistosis juxta ripas fluminis Cuanza gregaria, sed rarius florifera. X. STENOPHYLLA, Baker. Vellosia (Ophiothamna) stenophylla, Welw. Herb. Fruti- cosa, ramis 6-9 lin. crassus, basibus foliorum delapsorum rigidis adpressis crebre nervatis apice truncatis eroso-dentatis. Folia ad rosulam 6-8 rigidissima triquetra facie canaliculata 6-8 poll. longa 1-14 lin. lata glabra acuminata creberrime ner- vata dorso scabridula. Flores non vidi. /Мовватейев, in montosis petrosis arenoso- schistosis prope Cazemba, dense gregaria. Order H Y PoXIDACE Ж. Нүрох18, Linn. Н. ANGUSTIFOLIA, Lam. Pungo Andongo et Huilla, regio temperata, in herbidis arenosis. H. CUANZENSIS, Welw. Herb. Tuber perenne 12-2 poll. longum, 1 poll. crassum fibris setosis coronatum. Folia 8-10 linearia rigida parce pilosa distincte multinervata pedalia vel sesquipedalia 2 lin. lata. Scapi 2-4ni 3-4-роШсагев. Corymbus 3—4- florus, bracteis setaceis pilosis 5-6 lin. longis, pedicellis 2-3 lin. longis. Ovarium turbinatum 14 lin. longum, pilis brunneis vestitum. Ретїап її limbus luteus 3 lin. longus, segmentis exterioribus dorso pilosis. Pungo Andongo, in pratis humidis. . Floret Martio. Н. caANALICULATA, Baker. Tuber perenne oblongum 9-12 lin. crassum, fibris et squamis brunneis coronatum. Folia plura rigida erecta subteretia facie canaliculata 3—6 poll. longa 4 lin. lata, supra basin glabra contorta. Scapi 6-8 graciles albo-pilosi biflori 1-14 poll. longi. Pedicelli 2-3 lin. longi dense pilosi, bracteis minutis subulatis 266 MR. J. С. BAKER ON THE LILIACEZE ETC. зай. Ovarium turbinatum 2 lin. longum, pilis albis dense vestitum. Perian- thii limbus luteus 4 lin. longus, segmentis exterioribus dorso pilosis. Stamina perianthio duplo breviora. Capsula medio circumscissa. Ниша, regio subtempe- rata, in pascuis collinis arenosis prope Lopollo, frequens. Floret Dec.-Jan. Нүрохів MoNANTHOS, Baker. Tuber perenne oblongum 3 lin. crassum, fibris radicalibus earnosis. Folia 4-5 anguste linearia parce albo-pilosa 4—5 poll. longa, basi 2-1 lin. lata. Pedunculi uniflori 1-3ni 12-2 poll. longi parce pilosi gracillimi. Ovarium turbinatum 1 lin. longum subtiliter pilosum. Perianthii limbus luteus 2 lin. longus, segmentis exterioribus dorso viridulis subtiliter pilosis. Genitalia limbo duplo bre- viora. Capsula oblonga 3 lin. longa. Hwilla, regio temperata, in herbidis prope rivulum de Lopollo. Floret Dec. Н. ANGOLENSIS, Baker. Tuber perenne oblongum 3 poll. longum, 1-12 poll. crassum, setis densis ssepe coronatum. Folia 6-8 linearia rigida acuminata 6-9 poll. longa 2-3 lin. lata, ad margines et costam faciei inferioris solum albo-pilosa. Pedunculi 1-4ni 3-6 poll. longi, pilis albidis aseendentibus vestiti. Flores 6-8 in racemum 3-4 poll. longum dispositi, bracteis linearibus 3-6 lin. longis, pedicellis 3-6 lin. longis. Ovarium turbinatum 3 lin. longum, pilis densis albis ascendentibus vestitum. Perianthii limbus luteus 6 lin. longus, segmentis interioribus oblongis, exterioribus lanceolatis dorso dense pilosis. Stamina perianthio duplo breviora, antheris lanceo- latis sagittatis, filamentis deltoideis brevissimis. На, regio subtemperata, т col- linis dumetosis prope Lopollo, frequens. Floret Oct.-Nov. Н. РОГУЗТАСНУА, Welw. Herb. Tuber globosum vel depresso-globosum З poll. longum, basibus membranaceis brunneis foliorum delapsorum coronatum. Folia 10-12 rigida ensiformia pedalia vel sesquipedalia acuminata medio 8-9 lin. lata, facie glabra, dorso tenuiter pilosa. Pedunculi 6-8ni crassi 4-6 poll. longi. Racemus densus 20-30-florus 3-4-pollicaris, pedicellis 6-12 lin. longis, bracteis subulatis sericeis 6-9 lin. longis. Ovarium turbinatum 3 lin. longum breviter albo-pilosum. Peri- anthii limbus luteus 6 lin. longus, segmentis exterioribus dorso tenuiter pilosis. Stamina limbo duplo breviora, antheris basi sagittatis filamentis longioribus. Сар- sula sub apicem circumscissa. Пи а, im collinis dumetosis inter Lopollo et Ca- tumba. Floret Dec. Var. ANDONGENSIS, Baker. Folia angustiora 5-6 lin. lata, dorso dense albo-pilosa. Pedunculus compressus sesquipedalis. Pungo Andongo. CURCULIGO, Gaertn. C. GALLapaTENsIs, Schweinf. РІ. Callab. Exsic. по. 39. Tuber oblongum. perenne 2 _ poll. longum, 6-9 lin. crassum, fibris radicalibus carnosis. Folia produeta 3 lanceo- lata Е plicata utrinque tenuiter pilosa pedalia, medio 6-10 lin. lata. ОҒ WELWITSCH'S ANGOLAN HERBARIUM. 267 Flores radicales, singuli bracteis binis linearibus vel lanceolatis 15-18 lin. longis preediti, ovario piloso cylindrico, tubo supra ovarium filiformi 15-18 lin. longo, limbo luteo 4 lin. longo, segmentis lanceolatis, exterioribus dorso pilosis. Genitalia limbo triente breviora. Stylus complanatus apice stigmatoso-capitatus. Bacca oblonga 5-6 lin. longa. Golungo Alto, ай dumetorum margines аа Queta. Floret Nov. ; May prove a variety of Gethyllis pilosa, Schum. & Thonn. Pl. Guin. 172, of which we have specimens at Kew gathered by Barter. Order IRIDACE.E. GLADIOLUS, Linn. С. CHRULESCENS, Baker. Bulbus globosus 8-9 lin. longus, tunicis crassis verticaliter scissis. Caulis gracilis sesquipedalis. Folia 3-4 distantia superposita, infimum produetum lineare acuminatum rigidum glabrum 3-4 poll. longum 1 lin. latum, superiora rudimentaria brevissima. Spica laxa secunda 2-3-flora, spathe valvis oblongis obtusis 6-8 lin. longis viridibus margine membranaceis, interiore paulo minore. Perianthium sordide cærulescens, tubo curvato 5-6 lin. longo, limbo hori- zontali tubo szequilongo, segmentis superioribus oblongis obtusis 3—4 lin. latis, infe- rioribus angustioribus lanceolatis. Genitalia limbo subzequilonga, antheris 3 lin. longis. Huilla, regio subtemperata, in cultis anno elapso adhuc dumetis sylvestribus, ad Горойо. G. LURIDUS, Welw. Herb. Bulbus globosus 1 poll. crassus, fibris deorsum parallelis, superne reticulatis. Caulis gracilis pedalis. Folia basalia subcontigua 3-4 anguste linearia rigida glabra acuminata 1 lin. lata. Inflorescentia simplex vel ramosa. Spica secunda 8-7-роШсатів, spathz valvis lanceolatis viridibus purpureo tinctis 6-8 lin. longis, interiore paulo minore. Perianthii саге! tubus 5-6 lin. longus curvatus, limbo horizontali tubo equilongo, segmentis superioribus oblongo-spathu- latis obtusis 3—4 lin. latis, inferioribus magis unguiculatis. Genitalia limbo zequi- longa, antheris З lin. longis. иа, regio subtemperata, in arvis Zeæ Maydis prope Lopollo. “ Color florum exacte sicut in Iride lurida," Welw. G. smEviCAULIS, Baker. Bulbus globosus 1 poll. crassus, tunicis subtiliter fibrosis. Folia rosularum basalium 3 anguste linearia firma glabra semipedalia vel pedalia 3—4 lin. lata. Caulis 6-8-pollicaris, foliis paucis superpositis valde reductis solum preeditus. Spica ` simplex secunda 6-9-pollicaris 6—15-flora, sursum subdensa, spathee valvis lanceolatis acutis viridibus 6-9 lin. longis margine scariosis, interiore paulo minore. Perianthium violaceo-purpureum, tubo semipollicari, limbo hori- zontali pollicari, segmentis superioribus oblongo-spathulatis obtusis 5-6 lin. latis, inferioribus obovato-unguiculatis. Stamina perianthio triente breviora, antheris 268 MR. J. G. BAKER ON THE LILIACE# ETC. 4 lin. longis. Аиа, regio temperata, in pratis arenosis subhumidis inter Lopollo et Humputa, frequens. Floret Dec.-Jan. Habitus С. imbricati et С. palustris europei. GLADIOLUS LAXIFLORUS, Baker. Bulbus globosus 6-9 lin. crassus, tunicis brunneis mem- branaceis. Folia glabra linearia falcata rigide coriacea 6-15 poll. longa 8-4 lin. lata, nervis et marginibus stramineis incrassatis. Inflorescentia simplex vel furcata. Spica laxissima 4-8-роШсагін pauciflora, rachi flexuosa, internodiis inferioribus 12-2 poll. longis, spathze valvis 6-9 lin. longis purpureo tinctis acutis vel ает Perianthium læte roseo-purpureum, tubo curvato semipollicari, limbo роШсагі hori- zontali, segmentis oblongo-spathulatis obtusis subzequalibus 3—4 lin. latis. Genitalia . limbo paulo breviora. Capsula obovoidea 8-9 lin. longa. Ни а, in pratis paludosis alte herbidis, imprimis circa stagna juxta rivum de Lopollo. Floret Dec. Habitus etiam G.imbricati europei. G. cREGARIUS, Welw. Herb. Bulbus globosus 8-9 lin. crassus, tunicis subtiliter fibrosis contextis. Folia subbasalia rosulata 5-6 linearia glabra 6-9 poll. longa, 3 lin. lata. Caulis pedalis vel sesquipedalis. Spica simplex secunda erecta striata 5—8-pollicaris 10—12-flora, spathze valvis lanceolatis, exteriore pollicari vel sesquipollicari, interiore minore. Perianthium luteum, intus ad faucem maculis parvis 2 violaceis guttatum, tubo semipollieari curvato, limbi 9 lin. longi segmentis superioribus oblongo-spa- thulatis 4—41 lin. latis, inferioribus oblongo-unguieulatis. Genitalia limbo paulo breviora, antheris 4 lin. longis. Pungo Andongo, in rupestribus editioribus de Serra de Pedras de Guinga. Floret Jan.-Martium, “Crescit czspitosus, bulbis 4-10 dense gregatis.” Welw. б. BENGUELLENSIS, Baker. Bulbus globosus longe rhizomatosus, tunicis brunneis mem- branaceis. Folia basalia subrosulata anguste ensiformia dura glabra 6-9 poll. longa 6-8 lin. lata, nervis et marginibus crassis stramineis. Caulis pedalis et ultra, foliis 2-3 reductis preeditus. Spica laxa secunda 6-8-flora 6-8-pollicaris, spathæ valvis brunneis scariosis lanceolatis acutis 9-15 lin. longis. Perianthium splendide rubrum, fauce aurantiaco-purpureo maculatum, tubo curvato 8-9 lin. longo, limbi segmentis 9-12 lin. longis horizontalibus, segmentis superioribus ovatis acutis 6 lin. latis, inferioribus lanceolato-spathulatis flore expanso deflexis. Genitalia perianthio paulo breviora. Capsula (immatura) oblongo-clavata 8-9 lin. longa. “Ни а, regio {ет- perata, in pascuis dumetosis siccioribus prope Lopollo. Dec.—Feb. Bulbus rhizo- _ matosus more Lilii canadensis. | EU ‘Wetwitscutt, Baker. G. splendens, Welw. Herb., non Baker in Trimen Journ. Et 02016, 333. Bulbus globosus 15-18 lin. crassus, tunicis fibrosis. Folia rosularum _ basalium haud floriferarum linearia acuminata glabra rigide coriacea 9-15 poll. Е 8-4 lin. dato, nervis et marginibus crassis stramineis. Caulis semipedalis, ОЕ WELWITSCH'S ANGOLAN HERBARIUM. 269 foliis rudimentariis solum preeditus. Spica laxa secunda 6—10-pollicaris, spathze valvis lanceolatis acutis 13-2 poll. longis purpureo tinetis, interiore minore. Peri- anthium splendide rubro-aurantiacum, fauce maculatum, tubo curvato sesquipolli- cari, limbi bipollicaris segmentis superioribus horizontalibus ovatis acutis 10—12 lin. latis, inferioribus lanceolatis deflexis. Genitalia limbo paulo breviora, antheris 5-6 lin.longis. Capsula obovoideo-oblonga 15 lin. longa, seminibus oblongis late alatis. На, regio temperata, in collinis dumetosis argillaceo-arenosis circa Lopollo. Floret Oct.-Nov. GLADIOLUS MULTIFLORUS, Baker. Bulbus globosus 12-18 lin. crassus, tunicis fibroso-cris- patis. Folia basalia rosulato-disticha ensiformia modice firma 9-12 poll. longa 6-9 lin. lata, nervis vix incrassatis. Caulis pedalis. Spica densa stricta secunda pedalis, spathze valvis brunneis lanceolatis acutis 10-12 lin. longis. Perianthium rubro-purpureum, fauce pallidiore, tubo semipollicari abrupte curvato, limbo 9-10 lin. longo, segmentis superioribus brevioribus oblongo-spathulatis, inferioribus longioribus oblanceolato- unguiculatis. Genitalia limbo vix breviora, antheris 3 lin. longis. Capsula clavata 8-9 lin. longa. ИщИа, regio temperata, in sylvis, Protearum, loco dicto Monino, ecspitosis. Floret Feb. Ad G. spicatum, Klatt, accedens. С. ANGOLENSIS, Welw. Herb. Bulbus globosus 12-18 lin. erassus, tunicis fibris paral- . lelis preeditis. Folia superposita linearia vel ensiformia glabra rigide coriacea, infe- riora sesquipedalia vel bipedalia 6-9 lin. lata, nervis et marginibus stramineis incrassatis. Caulis 3—4-pedalis et ultra. Spica laxa secunda 6—18-pollicaris, spathze valvis brunneis lanceolatis acutis 13-2 poll. longis, interiore paulo minore. Peri- anthium luteum dense rubro maculatum, tubo curvato sesquipollicari, limbi hori- zontalis 18-21 lin. longi segmentis superioribus oblongo-spathulatis, superiore max- imo dorso convexo 12-15 lin. lato, inferioribus minoribus flore expanso decurvatis, omnibus obtusis. Genitalia limbo triente breviora. Capsula clavata 15-18 lin. longa ad basin trivalvis, valvis persistentibus divaricatis. Golungo Alto et Pungo Andongo, in graminosis frequens. Floret Feb. Ad С. psittacinum accedens. Preeterea legit Monteiro. G. quartinianus, A. Rich. Huilla, regio temperata, in pratis sylvaticis inter Lopollo et lacum Ivantella. Floret Dec. G. ANpoxGENsIS, Welw. Herb. Bulbus globosus 1 poll. crassus, tunicis brunneis fibris parallelis preditis. Folia ensiformia rigide coriacea acuta pedalia 10-12 lin. lata, nervis еб marginibus crassis stramineis. Caulis subpedalis. Spica laxa 6-8-роШ- caris, spathe valvis lanceolatis acutis brunneis 12-18 lin. longis. Perianthium aurantiacum, tubo curvato 15 lin. longo, limbi sesquipollicaris segmentis superioribus oblongo-spathulatis horizontalibus, superiore maximo dorso convexo 10—12 lin. lato, inferioribus minoribus decurvatis. Pungo Andongo, alt. 2400-3800 pedum, in rupes- tribus editioribus ipsius Presidii. Floret Dec. SECOND SERIES.—BOTANY, VOL. I. 29 270 MR. J. С. BAKER ON THE LILIACE# ETC. ANTHOLYZA, Linn. A. HUILLENsIS, Welw. Herb. Bulbus globosus 9-12 lin. crassus, tunicis pallide brunneis sursum membranaceis deorsum fibris parallelis preeditis. Caulis 6—15-pollicaris, foliis 3 superpositis lineari-subulatis, inferiore 3—4 poll. longo, preeditus. Spica 4-8- . pollicaris laxa secunda, spathe valvis brunneis obtusis 6-7 lin. longis. Perianthium splendide rubrum, tubo pollicari dimidio superiore cylindrico, limbi segmento supe- riore oblongo-spathulato 8-9 lin. longo, lateralibus superioribus oblongis 3 lin. longis, tribus inferioribus parvis lanceolatis. Genitalia segmento superiore paulo breviora. Capsula obovoideo-oblonga semipollicaris, seminibus discoideis distincte alatis. Ни” а, regio temperata, in dumetis arenoso-petrosis apricis inter Lopollo et Humpata. Ad A. Cunoniam accedens. ARISTEA, Soland. А. ANGOLENSIS, Baker. Folia basalia linearia distiche rosulata 6-9 poll. longa 3-41 lin. lata glabra crebre пегуаба. Caulis simplex anceps pedalis vel sesquipedalis, foliis pluribus basalibus similibus præditus. Inflorescentia pedalis anguste paniculata, floribus multis in capitulum confertis, capitulis terminalibus pedunculatis, latera- libus sessilibus, bracteis deltoideis membranaceis 3—4 lin. longis, exterioribus longi- oribus, dorso firmulis. Perianthii cærulei limbus cæruleus semipollicaris, segmentis post anthesin contortis. Capsula sessilis oblonga 3 lin. longa. Huilla, alt. 3800-5500 pedum, in pratis herbidis juxta ripas rivi de Lopollo. Floret Febr.-Maium. Monza, Linn. M. техтилз, Baker. Jridopsis textilis, Welw. Herb. Bulbus globosus 12-2 poll. latus, tunicis multis, fibris atris duris przeditis. Folium basale lineare solitarium glabrum rigidum distincte multinervatum 4—5-pedale 41-6 lin. latum. Caulis teres 2-3-ре- dalis simplex, foliis pluribus rudimentariis preeditus. Spathe 3-5-florz 4—5-polli- cares, valvis scariosis purpureis crebre nervatis, interiore majore. Pedicelli elongati demum exserti. Ovarium clavatum pollicare. Perianthii tubus supra ovarium | nullus; limbi saturate purpurei 23-3 poll. longi segmentis oblanceolato-spathulatis | . 5-6 lin. latis, exterioribus reflexis, interioribus erectis apice emarginatis. Stig- | mata 1} poll. longa, cristis lanceolatis 5-6 lin. longis. Filamenta 8-9 lin. longa, dimidio inferiore monadelpha. Ниша, alt. 3800-5500 pedum, in collinis dumetosis humidiusculis juxta rivum de Lopollo. Floret Aprili. Ad M. diversifoliam, Baker, abyssinicam accedens. Zridopsis, Welw. Herb. est stirps generis Moree floribus magnis persistentibus, caulibus robustis elatis, ovario clavato. м. Wixwrrscnn, Baker. Tuber globosum 10-12 lin. erassum, tunicis brunneis rigide fibrosis supra collum longe productis. Folia linearia acuta rigida pallide viridia саны distincte multinervata 11—21-pedalia, 5-6 lin. lata. Caulis pedalis simplex haud foliatus. әне terminales biflore, valvis acutis scariosis subzequalibus ОҒ WELWITSCH'S ANGOLAN HERBARIUM. 271 3—4 poll. longis. Pedicelli 2-4-pollieares. Ovarium clavatum pollicare. Perianthii tubus supra ovarium nullus, limbi saturate lilacini persistentis 2 poll. longi seg- mentis oblanceolato-spathulatis subzequalibus 5-6 lin. latis, exterioribus flore ex- panso faleatis, interioribus erectis. Stigmata 12-15 lin. longa, cristis lanceolatis 5-6 lin. longis. Filamenta 8-9 lin. longa, dimidio inferiore monadelpha, antheris 5-6 lin. longis. Capsula oblonga 15 lin. longa. — Hwilla, regio temperata in uligi- nosis et paludosis juxta rivum de ГороПо. Floret Nov. “ Irides quasdam europæas perfecte ludens," Welw. Мовжд GLUTINOSA, Baker. Tuber difforme 9-18 lin. crassum. Folia rosularum basalium haud floriferarum anguste ensiformia firma glabra pedalia 4—5 lin. lata. Caulis semipedalis, foliis 8—4 reductis preeditus. Inflorescentia pedalis, ramis 3—4 elongatis ascendentibus, infra nodos viscosis. Spathz valve oblongse obtuse virides arcte: complicate, interior longior 12-15 lin. longa. Pedicelli. spathse æquilongi. Peri- anthii tubus supra ovarium nullus; limbi pollicaris fugacis lilacino-purpurei seg- mentis oblanceolato-spathulatis subzqualibus. Filamenta semipolliearia infra medium connata, antheris flavis 1 lin. longis. Hwilla, regio subtemperata, in col- linis dumetosis prope Lopollo, cum Olematibus e£ Тіппеге sp. Floret Aprili. M. 8РїТНАМ ЖА, Baker. ‘Tuber globosum 3-6 lin. crassum. Caulis gracilis palmaris, basi foliis 2-3 anguste linearibus rudimentariis preeditus, dimidio superiore ramosus, ramis 2-3 infra nodos et spathas viscosis. Spathee 2-3 terminales pollicares triflore, valvis firmis viridulis striatis, exteriore minore semipollicari. Pedicelli spathze equilongi. Perianthii tubus supra ovarium nullus, limbi flavi fugacis 7-9 lin. longi intus maculis nigris decorati, segmentis oblanceolato-spathulatis subzqualibus. Filamenta 44 lin. longa prope basin connata, antheris oblongis 1 lin. longis. Hwilla, regio subtemperata, in dumetis arenosis circa Lopollo et Humpata. M. ANDONGENSIS, Baker. Tuber parvum globosum, fibris multis gracilibus præditum, Caulis semipedalis vel pedalis, dimidio superiore ramosus, infra nodos et spathas viscosus, foliis 2 anguste linearibus glabris rigide coriaceis 6-15 poll. longis 1 lin. latis praeditus, ramis paucis vel pluribus (2-8) arcuatis ascendentibus 1-3 poll. longis. Spathze terminales 6-9 lin. longi, valva exteriore 42-6 lin. longa. Pedicelli .Spathsz zequilongi. Ovarium turbinatum 1 lin. longum. Perianthii tubus supra ovarium nullus; limbi sulphureo-lutei fugacis semipollicaris segmentis oblanceolato- spathulatis subzequalibus. Filamenta 3 lin. longa supra medium connata, antheris oblongis 1 lin. longis. Capsula depresso-globosa 4 lin. lata. Pungo Andongo, alt. 9400-8800 pedum, in pratis sylvaticis paludosis breviter herbidis de Montollo. | - M. CaxNpELABRUM, Baker. Tuber parvum globosum. Caulis 2—3-pedalis, dimidio supe- riore pinnatim ramosus, ramis gracilibus erecto-patentibus infra nodos viscosis, inferioribus compositis, ramulis ultimis strictis erectis 1-3 poll. longis. Folia plura superposita acuta firma glabra linearia pedalia 3 lin. lata. Браге 4-5-flore, valvis 292 272 МВ. J. G: BAKER: ON THE LILIACEÆ ETC. valde insqualibus, interiore pollicari, exteriore obtusa 3-4 lin. longa. Pedicelli spathe sequilongi. Ovarium turbinatum 1 lin. longum. Perianthii tubus supra ovarium nullus; limbi lilacini fugacis segmentis oblanceolato-spathulatis sub- eequalibus. Stamina limbo dimidio breviora, filamentis dimidio inferiore connatis, antheris oblongis. Иша, regio subtemperata, in fruticetis petrosis ad Morro de -Lopollo, alt. 5200 pedum. Mor#A GRACILIS, Baker. Bulbi globosi ezespitosi 5-6 lin. lati tunicis subtiliter fibroso- reticulatis. Caulis strictus gracilis simplex 8-15-роШсагів, supra basin folio unico anguste lineari preeditus, infra nodos haud viscosus. Spathee solitaric terminales 1-6-flore, valvis lanceolatis 2-21 poll. longis, exteriore breviore. Pedicelli spathe demum equilongi. Ovarium clavatum 3 lin. longum. Perianthii tubus supra ovarium nullus; limbus luteus fugax 9 lin. longus, segmentis oblanceolato-spa- thulatis subzequalibus. Stamina limbo duplo breviora, filamentis dimidio inferiore connatis. Capsula oblonga 4-44 lin. longa. Huilla, im pascuis arenosis prope Lopollo, alt. 8800-5500 pedum. LAPEYROUSIA, Pourret. L. LITTORALIS, Baker. Caulis tripollicaris valde ramosus, foliis 2 linearibus acutis firmis glabris infra inflorescentiam instructus, rachi angulosa flexuosa, spathis plerisque pedunculatis terminalibus, 1-2 lateralibus sessilibus. Spatha valvæ semi- pollicares lanceolate viridule equales. Perianthium non vidi. Capsula mem- branacea obovoideo-oblonga 8-4 lin. longa. Mossamedes, in collinis sabulosis maritimis. | Г. ABYSSINICA, Baker. Montbretia abyssinica, В. Br. Pungo Andongo, alt. 2400- 3800 pedum. Г. CYANESCENS, Baker. Psilosiphon cyanescens, Welw. Herb. Bulbus globosus 1 poll. latus, tunicis fibris crassis preeditis. Caulis infra inflorescentiam 6-9-pollicaris, foliis 3-4 anguste linearibus firmis glabris pedalibus 2 lin. latis, nervis 5-7 crassis stramineis preeditis. Inflorescentia pedalis, rachi flexuosa ancipiti. Spathe valve lanceolatze sequales 15-18 lin. longe. Perianthii tubus filiformis 4—5-pollicaris, limbi albi violaceo tincti pollicaris segmentis oblongis 4-41 lin. latis. Genitalia limbo duplo breviora, antheris linearibus 4 lin. longis, filamentis brevissimis. Huilla, regio subtemperata, in collinis spongiosis editis prope Humpata. _ L. rRAGRANs, Baker. Psilosiphon fragrans, Welw. Herb. Bulbus ovoideus 8-9 lin. latus, tunicis crassis brunneis reticulato-fibrosis. Caulis brevissimus, foliis 2 line- aribus subcoriaceis acutis profunde nervatis glabris 6-8 poll. longis 2 lin. latis instructus. Inflorescentia laxissima corymbosa 6-9 poll. longa, rachi flexuosa angulosa. Spathz valve lanceolate subpollicares subsequales integre acute. Perianthium albidum, tubo cylindrico 4-42 poll. longo, segmentis lanceolatis acutis ОҒ WELWITSCH'S ANGOLAN HERBARIUM. 273 9-12 lin. longis. Anthere 44 lin. longæ, filamentis brevissimis. Capsula oblonga tubulosa 5-6 lin. longa. Huilla, regio subtemperata, in rupestribus et petrosis apricis prope Lopollo. LAPEYROUSIA ODORATISSIMA, Baker. (Plate XX XVI. figs. 2, 8.) Psilosiphon odoratissimus, Welw. Herb. Bulbus ovoideus basi applanatus, tunicis brunneis coriaceis reticu- latonervatis, collo 2-3-pollicari. Caulis 1—2-pollicaris, foliis 2-3 linearibus glabris erecto-patentibus 4-5 poll. longis 1 lin. latis, facie profunde canaliculatis. Spica solitaria densa multiflora. Spathæ valva exterior linearis viridis 8-5-роШсагіѕ, interior multo minor. Perianthii albidi fragrantis tubus cylindricus 4—5-pollicaris, limbi segmentis lanceolatis 12-15 lin. longis. Antherz ligulate З lin. longe, filamentis brevissimis. Zwille, regio subtemperata, іп dumetis arenosis petrosis prope Lopollo. DESCRIPTION OF PLATES. PLATE XXXIV. Figs. 1 to 3. Dipcadi comosum, Welw. 1. Plant, of natural size. 2. Single perfect flower from the lower part of the raceme, enlarged. 3. Unexpanded raceme. Figs. 4 to 7. Acrospira asphodeloides, Welw. 4. Of natural size. 5. Segment of perianth, enlarged. 6. Stamen ; and 7. Pistil in section, also enlarged. Pram XXXV. Figs. 1 to 6. Dasystachys campanulata, Baker. 1. Of naturalsize. 2. Segment of perianth ; 3. Stamen ; 4. Ovary : all magnified. Figs. 5 to 10. Dasystachys colubrina, Baker. 5. Plant, of natural size. 6. Segment of perianth; 7. Stamen; 8. Ovary; 9. A horizontal section of ovary ; 10. The seed: all magnified. Рглте XXXVI. Fig. 1. Xerophyta capillaris, Baker. Nat. size. Figs. 2 & 8. Lapeyrousia odoratissima, Baker. Of natural size. The above figures were drawn on stone, partly from sketches by Mrs. J. G. Baker, and partly from the objects themselves, kindly placed at the author's disposal by the Trustees of the late Dr. Welwitsch. Trans Linn.Soc. Ѕен 2 Bor. Voi] TaB. 34. MTS Baker del. +7 ACROSPIRA ASPHO! FB DIPCADI COMOSUM. MS Baker del. NN 27. EN УТ ) V TRAns Lom Soc. Sn 2.Bor Vou] Tan. 35 -o 1-4 РАЗУЗТАСНУ$ CAMPANULATA. 510 D.COLUBRINA. 2 20 JNFitch lith. / | А \ P = URES - АС 5 RIS АЛ PY ЧАД ANS я 3 | j М: Baker del. ` TRANS.LINN.Soc SER.2 Вот Vor..l TaB.36 JNFitch Hth. 1 XEROPHYTA CAPILLARIS. 2,3. LAPEYROU SIA ODORATIS SIMA ; 1:92; 2 XVIII. Contribution to the Lichenographia of New Zealand. By CHARLES ыб Esq., F.L.S., Auditor-General of New Zealand. (Plates XXXVIL, XXXVIII) Read March 1st, 1877. URCEOLARIA NOVJE-ZELANDLE, Kn. (Plate XXXVII. fig. 1.) Thallus cinerascens indeterminatus leproso-areolatus v. quasi detritus. Apothecia parva immersa, disco subdepresso fusco farinaceo (margine thallode nullo) nucleo ab exci- pulo dimidiato-fusco cincto, paraphysibus distinetis rectis tenuibus. Spore in ascis cylindraceis fuscze ovatze murali-divisze horizontaliter 5—6-septate longit. 027 mm., crassit. 018 mm. Ай saxa. РЕВТОЗАВТА GRAPHICA, Kn. (Plate XXXVII. fig. 2.) Thallus late effusus cartilagineus albicans v. cinereo-albicans rimuloso-areolatus, verrucis subgloboso-difformibus crebris v. confertissimis (deinde deplanatis) obsitus, gonidiis per totam partem instructus. Apothecia plura in singulis verrucis thalli inclusa (exeipulo proprio prorsus nullo) primitus a thallo tecta, tandem aperta, discis irregu- laribus заре іп pseudo-discum confluentibus, epithecio fusco, paraphysibus tenerri- mis tortilibus. Spore 4ne simplices ellipsoideze grumo-granulose dilute luteæ, epi- sporio crasso, longit. "082 mm., crassit. 04 mm. Ай saxa. LECIDEA LITTORALIS, Kn. (Plate XXXVII. fig. 3.) Thallus cinereo-albidus v. cinerascens crassus continuus у. areolatus late expansus levis. Apothecia adnata mujuscula (latit. 2:65 mm.) sparsa atra pruinosa tenuiter margi- nata dein tumidula et difformia, margine atro flexuoso v. lobato inciso. Spore simplices ellipsoideze, longit. “013 ad 018 mm., crassit. “006 mm. Hypothecium fus- cum v. linea atra hypothecii latera et basin circumscribens. Ad saxa. LECIDEA SUBGLOBULATA, Kn. (Plate XXXVII. fig. 4.) Thallus cinerascens minute furfuraceo-diffractus v. squamuloso-areolatus, granulis goni- mis viridibus. Apothecia immarginata interdum confluentia, disco atro convexo tandem hemisphserieo v. subglobuloso. Sporz in ascis ventricosiusculis simplices ellipsoideze subhyalinz tandem dilute luteze, longit. 013 mm. ad :022 mm., crassit. :007 mm. Hypothecium atrum v. fuscum. Ad saxa. LECIDEA SUBARGILLACEA, Kn. (Plate XXXVII. fig. 5.) Thallus luteo-albus continuus. Apothecia subaggregata innata v. adnata concava fusca immarginata interdum difformia вере furfuracea. Spore in ascis cylindraceis gru- moso-lutes v. hyalinze longit. -016 mm., crassit. "008 mm. — Hypothecium album. Ad terram argillaceam. Syn. Lecidea meiospora, Nyl. (?). 276 МВ. C. KNIGHT ОМ NEW-ZEALAND LICHENS. LECIDEA ATRO-MORIO, Kn. (Plate XXXVII. fig. 6.) Thallus nigro-cinerascens tenuis squamulosus, squamulis minutis rotundis planis rufe- scentibus nitidis ad ambitum plerumque minute furfuraceis. Apothecia atra parva plana ex hypothallo nigricante inter squamulas denudato oriunda, squamulas haud superantia, interdum margine thallino spurio granulis minutissimis cincta. Spore minutee ovoideze v. globosze, longit. “0065 mm., erassit. “005 mm. Gonidia magna. Hypothecium fuscum. Ай saxa. LECIDEA SUBLAPICIDA, Kn. (Plate XXXVII. fig. 7.) Thallus granulosus indeterminatus cinereo-albus. Apothecia in thalli lacunis szepissime confluentia adpressa aterrima nuda planiuscula v. concaviuscula, margine tenuissimo atro undulato, paraphysibus tenuibus conglutinatis. Sporæ in ascis ventricoso-clavatis parvulze ovoidez simplices hyaline longit. "005 mm., crassit. “004 mm. Stratum hypothecii subhymeniale atrum. Ad saxa. Differs from Г. lapicida іп its smaller ovoid spores, much finer paraphyses, and black hypothecium. In 7. lapicida the subhymenial stratum is slightly coloured, and the inferior stratum a thin black line continued from the margin. LECIDEA NIGRESCENS, Kn. Thallus nigro-cinerascens tenuis levis indeterminatus. Apothecia atra minuta convexa emarginata, paraphysibus conglutinatis. Spore in ascis ventricoso-clavatis minute ovoideze simplices hyaline longit. "004 mm., crassit. 008 mm. Stratum hypothecii subhymeniale atrum. Ad saxa. LECIDEA SCHISTACEA, Kn. (Plate XXXVII. fig. 8.) Thallus schistaceus determinatus tenuis squamulosus, squamulis rotundis adpressis minu- tissimis pruinosis hypothallo nigro. Apothecia nigra mediocria v. parva squamulas superantia marginata disco plano, margine prominente interdum undulato. Spore minute ellipsoidez incolores, longit. “008 mm., crassit. “004 mm. Hypothecium nigrum, Ad saxa. LECIDEA SUBCOARCTATA, Kn. (Plate XXXVII. fig. 9.) Thallus einereo-albus areolatus v. areolato-granulosus v. continuus. Apothecia conferta fusca v. atro-fusca adnata sat parva convexa, margine integerrimo pallido v. dilute fusco nudo demum evanido. Spore in ascis ventricoso-clavatis simplices ovoidex incolores, longit. “011 mm., crassit -007 mm. — Hypothecium album. Ad saxa. Differs from Z. coarctata in the much smaller ovoid spores, the ventricose ascus, and white есіне | LECIDEA SUBBADIO-ATRA, Kn. (Plate XXXVII. fig. 10.) Thallus fusco-cinereus rimulosus zequalis. Apothecia nigro-fusca convexa marginata, — dilute concolori in statu juvenili prominente, demum obscurato. Spore МВ. C. KNIGHT ОМ NEW-ZEALAND LICHENS. = 277 Ғаѕсе oblongs curvatule 1-septate, longit. “095 mm., crassit. “01 mm. Lamina prolifera ex hypothecio nigro duplici (strato intermedio fusco) enata. Ай saxa. Lecea УРнаклтірж, Kn. (Plate XXXVII. fig. 11.) Thallus albus parum visibilis v. obsolete indicatus. Apothecia parva atra, juniora plana et tenuiter marginata, dein convexa v. suborbiculata immarginata. Spore fuscee elliptieze longit. “013 mm., crassit. '006 mm. Hypothecium atrum. LECIDEA STELLULATA, Tayl. (?) Thallus albus v. cinereus tenuissimus minutissime areolatus v. granulosus hypothallo atro limitatus sepius in agellos ab hypothallo denudato metatus. Apothecia parva sub- innata confluentia atra plana tenuiter marginata, margine sspe e granulis effusis. Sporee in ascis ventricoso-clavatis ovate Газсге 1-верізіге, longit. "019 mm., crassit. “006 mm. Hypothecium atrofuscum. Ad saxa. LECIDEA SUBTUBULATA. (Plate XXXVII. fig. 12.) Thallus continuus fumosus tenuissimus. Apothecia elevato-sessilia tenuiter marginata, disco plano, margine atro elevato. Spore in ascis clavatis fusce plerumque in medio eonstrietze nonnunquam loculis tubulo brevi junctis, longit. 017 mm., crassit. "009 mm. Paraphyses distincte. Lamina proligera ex hypothecio stramineo enata ; linea atra hypothecii latera et in parte basin circumscribens. Ad saxa. LECIDEA PETRJEA, Flot., var. NEO-ZELANDICA, Kn. (Plate XXXVII. fig. 13.) Thallus cinereo-plumbeus minute granulosus, ex hypothallo nigricante inter granula denu- dato enatus. Apothecia sat parva plana crebra, disco aterrimo nudo elevato-margi- nato, margine primum e granulis minutis suffusis, tandem fusco-atro. Spore іп ascis ventricosis 3-5-septatze, et septa transversa septulis longitudinalibus v. obliquis juncta, subhyalinze tandem dilute fuscze, longit. "023 mm., crassit. “009 mm. Hypo- thecium fusco-nigrum. Ad saxa. L&crpEA PETRJEA, Flot., var. VIOLACEA. (Plate XXXVII. fig. 14.) Thallus minute tuberculosus, tubercula violacea, hypothallo nigricante. Apothecia parva ereberrima nigra innata v. adnata concaviuscula immarginata. Sporæ fuscze 5-septatee, septa transversa septulis longitudinalibus v. obliquis juncta, longit. "027 mm., crassit. 01 mm. Hypothecium atrum. LECIDEA TUBULATA, Kn. (Plate XXXVII. fig. 15.) Thallus granuloso-diffractus albo-cinerascens indeterminatus. Apothecia nigro margi- nata, disco plano pruinoso, margine atro elevato. Spore in ascis clavatis fuscze bi- loculares, loculis tubulo brevi junctis, longit. "018 mm., crassit. 011 mm. Нуро- thecium atrum. Hypothallus albus. Ай saxa. LECIDEA SUBFARINOSA, Kn. (Plate XXXVII. fig. 16.) Thallus albus effusus rugulosus. Apothecia (latit. 8 mm. v. minora) albo-pruinosa rotun- SECOND SERIES.—BOTANY, VOL. I. 2B 278 MR. С. KNIGHT ОМ NEW-ZEALAND LICHENS. dato-difformia elevato-sessilia tenuiter marginata, margine ѕере flexuoso, paraphysi- bus superne fuscescentibus concretis, disco concavo vel plano vel convexo. Spore 4-6-septatæ aciculari-fusiformes incolores vel tandem dilute lutescentes, longit. 084 mm., crassit. '005 mm. Ad cortices arborum. L. farinosa, Ach., L. Dillemana, Ach., L. abietina, Ach., Г. premnea, Ach., and the above-described Lichen are very closely allied. The New-Zealand Lichen approaches closely L. abietina, Ach., from which it differs principally in the greater number of septa, in the spores, and the smaller apothecia. TIHLOTREMA SAXATILE, Kn. (Plate XXXVIT. fig. 17.) Thallus albidus v. cinereo-albidus continuus granulosus indeterminatus tenuis. Apo thecia normaliter globosa in verrucis thallinis rotundatis supra planis immersa, aper- turis rotundatis depressis, excipulo proprio instructa, paraphysibus capillaribus. Ascus monosporus. Sporve dilute fuscee tandem fusco-nigricantes fusiformes murali- divise, longit. “16 mm., crassit. "038 mm. Ad saxa. THELOTREMA MONOSPORUM, var. PATULUM, Kn. (Plate XXXVII. fig. 18.) Thallus luteo-albus rimoso-diffractus erassus. Apothecia verruceeformia excipulo dupliei instructa, interiore membranaceo tandem lacero-dehiscente, nucleo madefacto expanso discoideo. Ascus monosporus interdum disporus (unde sporis minoribus). Spore fuscee fusiformes murali-divise, longit. “05 ad `1 mm., crassit. “015 ad ‘03 mm. Аа cortices arborum. Syn. Thelotrema monosporum, var. patulum, Nyl. in litt. ASTROTHELIUM PYRENASTROIDES, Kn. (Plate XX XVII. fig. 19.) Thallus effusus glauco-ochraceus у. cinereo-albescens у. olivaceus rimulosus. Apothecia сопуеха a thallo plus minus velata demum denudata carbonacea, loculis 2—5, ostiolis convergentibus in os commune saepissime desinentibus, ore ochraceo, paraphysibus tenerrimis rectis v. subtortilibus. рога in ascis elongato-cylindraceis incolores de- mum dilute fuscze fusiformes 5-8-septatæ (vix unquam 3-4-septatw) v. sepius murali-divis:e v. interdum locellis medianis a septulis longitudinalibus divisis, longit. *038 mm., crassit. 012 mm. Ай cortices arborum. Syn. Verrucaria pyrenastroides, Kn., Trans. Linn. Soc. vol. xxiii. p. 100, tab. ii. ; Astrothelium prostratum, Stirton, Journ. Linn. Soc. vol. xiv. p. 473 ; A. ochrocleistum, Nyl., in litt. An Trypethelium pyrenuloides, Mont.?, T. Cumingii, Fee?, T. astroidea, Fee?, Verrucaria _ Thwaitsii, Leight. ? EC m GEMELLIPARA, Kn. (Plate XXXVII. fig. 20.) Thallus flavo-zeneus v. flavo-olivaceus tenuissimus glaber linea nigra limitatus. Apo- thecia parva parum prominula, ambitu applanato, dimidiatim nigra, basi amplificata, о pe е uso ее p distinctis. Бровр obovate uniseptate ssepe in MR. C. KNIGHT ОМ NEW-ZEALAND LICHENS. 279 medio subconstrictee cellula superiore majore incolores, longit. '017 mm., crassit. '006 mm. Ad cortices arborum. Syn. Verrucaria circumpressa, Nyl., in litt. ; V. epidermidis, var. gemellipara, Kn. VERRUCARIA MINUTISSIMA, Kn. (Plate XXXVII. fig. 21.) Thallus atro-cinerascens tenuissimus. Apothecia contigua minutissima prominula dimi- diatim nigra poro pertuso instructa, paraphysibus distinctis. Spore dilute fusce uniseptate (an interdum 3-septatæ?) in medio subconstrietz, cellula superiore majore, longit. “018 mm., crassit. “005 mm. Ай cortices arborum. The spores have much the appearance of а Spheria; exeept the bright colour, not unlike those of Verrucaria conferta, Tayl. (Leight. Angioc. Lich.). VERRUCARIA SUBBIFORMIS, Kn. (Plate XXXVII. fig. 22.) Thallus albus effusus ineequalis rimosus. Apothecia subminuta prominentia (madefacta innata) nigra subglobosa integra, paraphysibus capillaribus confertis. Spore in ascis cylindraceis oblong 1-septatze incolores, longit. `02 mm., crassit. "007 mm. Ad cortices arborum. BaccLrETTOA, Mass. (?), OCELLATA, Kn. (Plate XXXVII. fig. 23.) Thallus fusco-cinereus continuus tenuis subglaber, gonidia magna. Apothecia in verrucis prominulis immersa interdum 2—4 confluentia, excipulo proprio carbonaceo instructa, primo punctiformia dein aperta, margine thallino albo sculpto-angulari v. seulpto- subrotundo v. irregulariter dehiscente. Nucleus globosus verrucoideus, amphithecio grumoso oriundus, paraphysibus crassis mucoso-diffluxis farctus. Spore in ascis curvatulis fusiformibus ellipsoideze incolores v. luteo-grumoss, longit. "028 mm., crassit. 012 mm. Ай saxa. VERRUCARIA OCCULTA, Kn. (Plate XXXVII. fig. 24.) Thallus crustaceus levigatus aureus v. ochraceus v. ochraceo-albus verrucoso-colliculosus irregulariter plicato-rugulosus plus minus rimosus. Apothecia numerosissima glo- bosa in verrucis thalli confluentibus penitus abdita, ab excipulo simplici proprio car- bonaceo clausa, canaliculis ad sporas mittendas in verrucas thalli productis, ostiolis paululum depressis minutis instructa, paraphysibus distinctis capillaribus. Spore in ascis elongato-cylindraceis oblongo-ovoideze 3-septate dilute fusce v. subincolores longit. 016 mm., crassit. “006 mm. Ай cortices arborum. Ап У. micromma, Mont. ? VERRUCARIA PRUINO-GRISEA, Kn. (Plate XXXVII. fig. 25.) Thallus effusus tenuis plus minus pseudo-farinaceus griseus. Apothecia minutissima dimidiata patentia. Spore in ascis clavatis incolores tandem fuscæ lineari-oblongze l-septatze v. seepissime interrupte quadrinucleolatze, longit. "02 mm., crassit. “006 mm. Ad cortices arborum. 2в2 280 MR. C. KNIGHT ОМ NEW-ZEALAND LICHENS. VERRUCARIA DEALBATA, Kn. (Plate XX XVIII. fig. 1.) Thallus albescens tenuis effusus glaber. Apothecia nigra parva hemisphzrico-conoidea dimidiata, paraphysibus distinctis. Spore in ascis cylindraceis dilute fusce ellip- soidez 8—5-septatee, longit. 023 mm., crassit. 008 mm. Ad cortices arborum. VERRUCARIA SAXICOLA, Kn. (Plate XXXVIII. fig. 2.) Thallus luteo-olivaceus tenuissimus v. nullus. Apothecia parvula hemispheerica, peri- . thecio carbonaceo integro, ostiolo inconspicuo, nucleo subgloboso hyalino, paraphysi- bus capillaribus farcto. роге іп ascis elongato-cylindraceis incolores fusiformes 7—8-septatze, longit. 032 mm., crassit. 005 mm. Ad saxa. VERRUCARIA ASTATA, Kn. (Plate XXXVIII. fig. 3.) Thallus tenuis fulvo-fuscus v. cinereo-fuscus opacus continuus indeterminatus. Аро- thecia nigra prominula (madefacta szepe innata) hemispheerica integra, paraphysibus .eapillaribus tenerrimis distinctis. Spore in ascis elongato-cylindraceis ovoidez fuseze 3-septatze, longit. '04 mm., crassit. "009 mm. Ай cortices arborum. Probably this may be У. aspistea, Ach.; but there is much confusion about Acharius's plant. Nylander remarks “Esse videtur modo V. nitida minor, sporis minoribus." Montagne describes У. aspistea, Ach., “sporis magnis ellipticis 16-annulatis, annulis pauci-cellulosis;" he adds that V. aspistea of Eschweiler is incorrectly named, and he has therefore named Eschweiler’s plant V. Eschweileri with “spora binucleolata." Acharius describes his Lichen as having a polished yellowish thallus, limited by a black line, in these and other characters differing from the New-Zealand Lichen. PORINA ENDOCHRYSA, Mont. (Plate XX XVIII. fig. 4.) Thallus late effusus tenuis fragilis colliculosis glauco-cinereus subtus bullatus. Apothecia immersa, perithecio flavescente integro normaliter globoso, ostiolo fuseo depresso primum occluso demum aperto, nucleo fusco, paraphysibus filiformibus. Spore 8ne incolores eymbiformes murali-divise, longit. 09 ad '15 mm., crassit. "025 ad “04 mm., episporio crasso. Ad cortices arborum. Syn. Thelenella wellingtonii, Stirton, Journ. Linn. Soc. vol. xiv. p. 473. CHIODECTON (PLATYGRAPHA) INCONSPICUUM, Kn. & Mitt. (Plate XXXVIII. fig. 5.) Thallus crustaceus cinereo-violaceus verruczeformis, verrucis dilute concoloribus. Apo- thecia minuta normaliter globosa, in verrucis thallinis rotundatis v. flexuoso-elon- gatis supra planiusculis immersa (passim confluentia) plura in quavis verruca unum- . . quidque ex atro-fusco toro oriens, disco rotundo v. oblongo v. attenuato-oblongo . nigro. роге oblongo-fusiformes curvatule 3-septatz incolores, longit. "00 mm., E _ erassit. "004 mm. Ad cortices arborum. уп. Chiodecton conchyliatum, Stirton; С. moniliatum, Stirt.; С. sinuosum, Stirt. Ореке: SAXICOLA, Ach. (Plate XXXVIII. fig. 6.) Thallus obliteratus, Apothecia atra lirellzeformia заере flexuosa rarius rotundo-difformia, MR. C. KNIGHT ON NEW-ZEALAND LICHENS. 281 epithecio pliciformi v. rimeeformi, marginibus parallelis v. medio paullulum distentis. Spore lineari-oblongze v. lineari-elavatse 3-septate dilute lutex, longit. “017 mm., crassit. 006 mm. Ай saxa. FrssURINA Nov -ZELANDLE, Kn. (Plate XX XVIII. fig. 7.) Thallus crustaceus lzvigatus ochraceus tenuiter areolatus. Apothecia immersa flexuosa lineari-elongata, fissuris а thallo marginatis conniventibus tumidulis, excipulo atro- fusco crasso a thallo insuper tecto infra subito evanescente. Spore in ascis clavatis 5-septatz ellipsoidez incolores, longit. 02 mm., crassit. “01 mm. Ай saxa. MELASPILEA METABOLA, Nyl. (Lich. Nov: Caledonism). (Plate XXXVIII. fig. 8.) Thallus cinereo-albidus v. cinereus diffracto-rimosus v. areolatus, tenuissimus. Apothecia nigra у. fusca parva rotundato-lecideoidea (juniora punctiformia) convexa immargi- nata, intus pallido-incoloria. Spore 2-Sne oblongo-cylindracew inferne magis attenuatz et сигуай а (52ере in capillam product) murali-divise v. multiloculares, longit. 0:075 ad 0-1 mm., crassit. 0:01 mm. Thecw pyriformes infra іп caudam non- nihil flexuosam product. Hypothecium dilute fuscescens. Syn. Melaspilea amphorodes, Stirton. PHLYCTIS SORDIDA, sp. nov. (Plate XXXVITI. fig. 9.) Thallus cinereus fragilis tenuis v. crassus contiguus subgranulosus inzequalis у. corrugato- rimosus v. minutissime areolatus. Apothecia minuta rotundato-difformia, margine thallino albino-leproso cincta, disco fusco nonnihil farinoso. Spore 5-—8-septatze ellipsoidez curvatz incolores, longit. 0:06 mm., crassit. 0-007 mm. Ad corticem. Obs. When the thallus is thin it necessarily adapts itself to the surface on which it grows, and may thus be either smooth, rugose, rimose, &c. PHLYCTIS OCELLATA, sp. nov. (Plate XXXVIII. fig. 10.) Thallus albido-canescens v. dilute murinus fragilis corrugato-rimosus. Apothecia minuta regularia in tuberculo innata, a thallo marginata, nucleo strato gonimo imposito, disco fusco concaviusculo. Spore ellipsoidee incolores 3-6-septate, longit. 0028 mm., crassit. 0-004 mm. Ad corticem. Рніүстіз NEO-ZELANDLE, sp. nov. (Plate XXXVIII. figs. 11 & 12.) Thallus albus crassus v. tenuis rimosus fragilis levis v. granulosus. Apothecia minuta solitaria v. aggregata innata rotundato-difformia a thallo pruinoso velata vel margine thallino albido-leproso cincta, disco dilute fusco albido-pruinoso. Spore in ascis obovato-clavatis normaliter 8nz curvate cylindraceo-ellipsoide:, apicibus acutis v. obtusis dilute flavescentes 6-9-septatz, longit. 0:07 mm., crassit. 0-006 mm. Para- physes distincts. Ad corticem. Obs. There is a variety with thin white granular thallus; spores cylindrical, rounded at either apex. 282 МВ: C. KNIGHT ОМ NEW-ZEALAND LICHENS. STICTA CANALICULATA, Sp. nov. Thallus glauco-pallescens late expansus (latit. usque pedalis) rigens vix nitidiusculus levis lineari-laciniatus canaliculatis, laciniis insignius sinuato-pinnatifidis, subtus totus tomentosus luteo-fuscescens, cyphellee albidze suburceolatze mediocres v. majuscule. Apothecia badio-rufa v. fusca (latit. 5 mm.) margine crenulato, receptaculo granulato. Sporæ leviter fuscescentes fusiformi-oblonge uniseptatse, longit. 0025 mm., crassit. 0:01 mm. Obs. Much stouter than S. variabilis, and differs from S. damecornis in the white larger cyphelle, the channelled laciniz, and the under surface entirely covered by a brown tomentum. PLACODIUM ILLITUM, sp. nov. (Plate XXXVIII. fig. 13.) Thallus contiguus determinatus fusco-cinerascens v. sordide albus arcte adnatus levis v. in centro minute granulosus in ambitu non radioso-sublobatus sed minutissime crenatus gonimicus вере cephalodiis fusco-luteis radiatim onustus. Apothecia crebra v. conferta unde per mutuam pressionem tandem difformia, margine cinerascente v. albo subinde evanescente, disco fusco-nigrescente concavo per setatem plano et mar- ginem subeequante. Paraphyses discretæ. Spore in ascis elongato-cylindraceis uniserialiter octonæ lutes ellipsoideze, longit. “0192 mm., crassit. '007 mm. Hypo- thallus fuscus. АЯ lapides. Differs from P. gelidwm, Ach., in the darker non-lobate thallus, in the colour of the disk, and the thinner margin of the apothecia, and the dark brown hypothallus. Р. Rhuderi has an albido-lutescent and lobate thallus, the disk of the apothecia atro-pruinose, and the spores not exceeding “008 mm. in length. PLACODIUM ARGILLACEUM, sp. nov. (Plate XXXVIII. fig. 14.) Thallus cinereo-albus farinaceus polline albo obsitus continuus indeterminatus. Cephalo- dia carnea centrali radiatim rimosa. Apothecia urceolata disco carneo, margine crasso albo farinaceo. Spore in ascis elongato-cylindraceis uniserialiter octonze ellipsoideze, longit. "016 mm., crassit. "008 mm. Supra terram. | PLACODIUM LECANORINUM, sp. nov. (Plate XXXVIII. fig. 15.) Thallus albido-cinerascens continuus determinatus arcte adnatus tenuis levis v. subgranu- lato-pulverascens. Cephalodia copiosa fusca v. fusco-atra madefacta indistincte radiatim rugosa. Apothecia sordide carnea concava, margine albido interdum fari- пасео, paraphysibus distinctis. Spors іп ascis elongato-clavatis obovate 3-septate, longit. “018 mm., crassit. "006 mm., hyaline v. pallide luteæ. — Zypothallus pallide fulvus. Ad lapides. MR. C. KNIGHT ОМ NEW-ZEALAND LICHENS. ; 288 ‘EXPLANATION OF THE PLATES. [The whole of the figures in these Plates have been greatly magnified ; but in his MS. the author has omitted to add the degree of enlargement of each object.—Epzrror. | PLATE XXXVII. . Spores of Urceolaria Nove-Zelandie. . Ascus of Pertusaria graphica. . Ascus of Lecidea littoralis, . Г. subglobulata. а. Ascus. with spores; $. Squamule with apothecia. Fig. 5. L. subargillacea. а. Section of apothe- cium ; 0. Its ascus. 6. Ascus of L. atro-morio. 7. Ascus of L. sublapicida. Fig. 8. Ascus of L. schistacea. 9. Ascus of L. subcoarctata. Fig. 10. L. subbadio-atra. а. Ascus; б. А sepa- rate spore. i "Ow me он Е». 11. Г. Whakatipe. а. Ascus; b. A single spore. Fig. 12. L.subtubulata. a. Ascus; b. Spores in different stages. Fig. 13. Spores of L. petrea, var. Neo-Zelandica, in different stages. Fig. 14. Spores of L. petrea, var. violacea. Fig. 15. L. tubulata. a. Ascus; b. Separate spores. Fig. 16. Spores of L. subfarinosa. Fig. 17. A spore of Thelotrema sazatile. Fig. 18. A spore of T. monospermum, var. patu- lum. Fig. 19. Astrothelium pyrenastroides. а. Section of apothecium, showing at os the ostioles of adjoining cells opening into common canal; б. A spore of same lichen. Fig. 20. Verrucaria gemellipara. а. Apothecium in section; 6. A spore. Fig. 21. V. minutissima. а. Apothecium in sec- tion; 0. Spores of same. Fig. 22. V. subbiformis. а. Section of apothe- cium ; б. Spores. Fig. 23. Baggliettoa, Mass.?, ocellata, Kn. а. Apothecium in section; 0. Ascus of same lichen. i Fig. 24. Verrucaria occulta. a, b, c. Various sections of apothecia; d. Spores of same lichen. Fig. 25. V. pruino-grisea. a. Section of apothe- cia; 6. An ascus; с. Spores in diffe- rent stages. Ртлте XXXVIII. Fig. 1. Verrucaria dealbata. a. Apothecium in section; 6. Spores. Fig. 2. V. saxicola. a. Apothecium in section ; b. Ascus; с. А spore. Fig. 3. V. astata. a, b. Sections of apothecia ; c. Spores. Fig. 4. Porina endochrysa. a. Apothecium in section ; б. Two spores. . 5. Chiodecton (Platygrapha) inconspicuum. а. Vertical section, displaying several apo- thecia ; 5. Horizontal section of same ; c. A septate spore. ig. 6. Opegrapha saxicola, a. Section of apo- thecium ; б. Ascus; с. Spore. .7. Fissurina Nove-Zelandie. а. Apothe- cium in section; 0. Portion of ascus. Fig. 8. Melaspilea metabola. a. Section of apo- thecium; 0. Ascus; c. Upper portion of ascus. Fig. 9. PAlyctis sordida, its separate spores. Fig. 10. P. ocellata. а. Two apothecia in section. b. Ascus; c. Five separated spores. Fig. 11. P. Neo-Zelandie. а. Ascus; b. А para- physis. Fig. 12. P. Neo-Zelandie, var. Septate spore of same. Fig. 13. Ascus of P/acodium illitum. Fig. 14. Ascus of P. argillaceum. Fig. 15. Ascus of P. lecanorinum, а. Ап ascus ; б. Trans. Шімм.Зос Ѕев.2 Вот VoL I TA 37 C Knight. NEW-ZEALAND LICHENS. DBlair lith. D Blair lith. Trans. Linn.Soc.Ser.2 Bor Vou I Tas 38. <-> $e ка =, loss er > 5252 LI - s, - RAE = 2000000696727 CIS ез» бтр ет» NEW-ZEALAND LICHENS. [ 285 1 XIX. On some Points т the Morphology of the Primulacee. Ву M. T. MasrEns, M.D., Ғ.Е,ғ., F.L.S., $c. (Plates XXXIX.-XLI.) Read June 7th, 1877. THE most generally interesting points connected with the structure of the flowers of the Primulace: reside in the superposition of the stamens to the petals, the free central pla- centa, and the nature of the ovules. Much discussion has taken place with reference to all these points. The explanation of the observed facts, and the arguments and inferences based upon them, have been de- rived from a study— 1. Of the comparative morphology of the genera of the order and of its allies. 2. Of the mode of development or organogenesis of the parts of the flower. 3. Of the minute anatomy of the flowers, and especially of the distribution of the vascular bundles within them. 4. Of the teratological phenomena observed in the order. These latter are exceedingly numerous, and have consequently attracted the attention of a proportionate number of observers. It is to them that I propose to devote much attention in this paper, though I would deprecate the notion that teratological changes per зе are to be taken as safe guides to the explanation of structure. The evidence they afford always requires to be controlled by that derived from other sources. I must preface my remarks by stating that I have personally examined during the last . twenty years a very large number of such instances, while my note-book teems with references to what has been published by others in this country and abroad. It would be hardly possible to cite all these references, and certainly not desirable to do so; for many of them are, relatively speaking, of little importance to the matters I propose now to bring under the consideration of the Society. Such, for instance, are the malformations affecting the stem, leaves, or inflorescence, phenomena interesting in themselves, but which throw no special light on the structure of the order. In passing, I may make allusion to the very great frequency with which deviations from the usual or, as it is called, typical structure occur in this group. Ав is well known to all who pay attention to these matters, malformations are much more common in some orders than in others. In the case of Primulacez a glance at the recorded literature will suffice to show how unstable, if I may so say, is the structure of certain genera of this order. Were it possible to enumerate individual cases, the fact would be even more prominently noticeable. From the frequency with which thas changes occur in such genera as Primula and Cyclamen it might be inferred that the changes induced by cultivation afforded sufficient reason for these structural aberrations. No doubt this is to a certain extent true. But SECOND SERIES.—BOTANY, VOL. I. 2s 286 DR. M. T. MASTERS ОМ SOME POINTS IN even among wild plants not subjected to cultivation the relative proportion of vari- ously deformed Primulacez is decidedly large. The common Primrose is, for instance, very subject to such changes, while the species of Lysimachia and Anagallis, which are comparatively rare in gardens, are so in a less degree. | Particular genera and even particular species are much more subject to these changes than others—the wild Primrose, for instance, as compared with the wild Cowslip. So in cultivation Primula prenitens (P. sinensis) is exceedingly liable to morphological disturbances, while Primula cortusoides and P. japonica are at present, according to my experience, but very slightly so. It may be said, perhaps, that the varying conditions under which wild plants of widely extended geographical distribution grow may have some effect in determining these changes; but if so, they are not very obvious; for one may find a single patch of mal- formed Primroses growing amid a multitude of normal flowers, and apparently under like conditions. Samolus Valerandi has almost world-wide distribution; but it grows every- where in similar localities ; hence it is not surprising to find that, comparatively speaking, this species is not much subject to monstrous deviations from the typical structure. It _ is worthy of consideration whether this instability of form in certain species is connected with any change, either in the way of advance or retrogression—whether, in other words, it is in any way connected with the progressive development of an advancing type, or the gradual degeneration of a form destined to more or less speedy extinction. This, how- ever, is a question upon which it is not my purpose to enter in this place. I propose now to glance briefly at some of the more important changes which are met with in the plants cf this order. Catyx.—The most frequent deviations from the normal type of calyx consist іп an enhanced degree of leafy development, or phyllody—a very common occurrence in the Chinese Primrose of gardens (Primula prenitens), and by no means uncommon in wild - Primroses. | Sometimes the leafy character is associated with complete dialysis of the sepals, when there results a calyx of five free leaves, sometimes distinctly petiolate, as in ће Cycla- men figured on the next page*. A minor degree of this change, that called * virescence,” is also not uncommon. Here the parts of the flower are nearly unchanged in form and position, but are of a green colour. | А change of an opposite character is that wherein the calyx assumes the coloration. of the corolla, the condition called * calycanthemy.” This occurs in some long-cultivated varieties of Polyanthus, and has been made the subject of a special memoir by the late Prof. Morren 7. . Conorra.—The teratological changes which occur in the corolla of Primulacee are of special interest in connexion with the notions entertained by morphologists as to its con- stitution. In most of the genera of the order the five petals of which the corolla consists form, at the base, a tube of varying length and diverse form, while above the five * Masters, * Vegetable Teratology,’ p. 248. + Bull. Acad. Belg. xix. part 2, p. 93. THE MORPHOLOGY OF THE PRIMULACEZ. 728 segments are free. Attached to the corolla-tube, and superposed to its segments, are the five stamens. In Glawx the corolla is wanting. Іп Samolus five scales emerge from the throat of the corolla, alternating with the stamens, which naturally suggests the notion that they represent an outer row of stamens, which in most cases is suppressed*. In a double- flowered Cyclament lately examined by me there was a double row of petals, the inner occupying the position of the scales of Samolus. Others consider that the petal and the stamen are in this case not two distinct parts but two divisions of the same part. Thus Duchartre? describes the petals as appearing after the stamens as уњ. о а sort of appendage or dédoublement:—“Elle (the GEBEN corolla) ne constitue en effet dans l'origine qu'un bour- 2 | relet ou un petit repli qui entoure la base du verticille staminal" Van Tieghem $, basing his opinion on tlie faet that a common fibro-vascular bundle divides into an outer and an inner division, the former for the petal, the latter for the stamen, comes to the conclusion that stamen and petal form part of the same phyllome. А similar conclusion is arrived at by Pfeffer|| from the study of the development of the flower. He shows that within the calyx five small tubercles are deve- loped, alternating with the sepals. Hach of these tu- bercles ultimately develops into a stamen; and from the outer surface of this stamen is gradually protruded the petal. According to this view, then, the Primrose has, morphologically, no corolla, the petals being merely outgrowths from the stamens €. Development of the Parts of the Flower.—M y own observations on the development of the flowers of Primula and Lysimachia lead me to a slightly different conclusion. In the first place, the mode of development does not appear to be uniform in the same genus; thus in some specimens of Lysimachia nummularia it has appeared to me that the petals did really sometimes (but not always) precede the stamens in their develop- ment. When this happens the latter advance so much more rapidly that they speedily outstrip the petals, and appear as if they had preceded them. If this be really so, the petal, instead of being an outgrowth from the stamen, is, in such cases, the organ from which the latter proceeds (Plate XXXIX. figs. 10-23). Many double Primroses, in which petal succeeds petal, one in front of another, the inner bearing a stamen, support this view. - In normal Primroses the course of development, according to my observation, is as Cyclamen with leafy calyx, referred to in the text. + *Gardener's Chronicle, March 17, 1877 * Richler, Bliithen-Diagramme, p. 327. $ Ann. Se. Nat. 1867, ix. р. 135. + Ann. Se. Nat. 1844, 1. р. 279. | Pringsheim, Jahrbuch für wissensch. Botan. 1871, viii. р. 194. 4 Frank (in Jahrbuch, 1875, t. x. p. 204) also says that the petalline tubercles are at first not distinct from the stamens. 252 288 DR. M. T. MASTERS ОМ SOME POINTS ІМ follows :—The rounded end of the receptacle gives off first of all five sepals in successive order. These attain relatively considerable dimensions before the five staminal tubercles which come next in order are simultaneously protruded (Pl. XX XIX. fig. 1) and alter- nately with the sepals (Pl. XXXIX. fig. 2). After the staminal tubercles have advanced somewhat—after, indeed, the antheral portion is separated from the filament, the tubercles which are to form the petals are protruded apparently from the back of the filament (Pl. XXXIX. figs. 8,4). Isay apparently, because what really takes place (in Primula) is this. After the formation of the staminal tubercles the thalamus pushes out the petalline tubercles, one at the -very base of the outer surface of each of the filaments, and so close to it that it seems to arise from the filament itself; the receptacle, continuing to grow in the form of a tube (Pl. XXXIX. fig. 4), upraises the corolla- tube, and with it the stamens. While thisis taking place the extreme apex of the recep- tacle is pentagonal and tabular in appearance, or slightly depressed in the centre (PL XXXIX. fig. 2). On its raised edge five small tubercles sometimes appear simul- taneously, which are the carpels (Pl. XXXIX. fig. 5). These five tubercles are gra- dually raised from below by the lengthening of the receptacle in a tubular manner (Pl. XXXIX. fig. 6), so that we have soon a flask-shaped pistil with a narrow neck and open mouth (figs. 7 & 8). Within the rudimentary pistil so sketched out the receptacle changes its former flat or depressed condition for a convex or dome-shaped appearance; the middle portion of this dome speedily becomes covered with ovules from above down- wards; the upper portion elongates into a naked cone; and the lower portion forms a stipes, which is apparently quite free from the walls of the pistil. Fig. 9 shows a plan of a vertical section through the flower at this stage. "The neck of the pistil gradually lengthens into a style, becomes closed; and so we have ultimately the perfect ovary, with whose appearance every one is familiar *. Chorisis.—Adverting now to the teratological appearances presented by the individual petals, it may be mentioned that the petals of Primulacez are frequently the subjects of lateral chorisis or fission. Slight instances of this are seen in the petals of the common Primrose, in the central notch by which they are characterized. If this subdivision be carried further we may have deeply bipartite petals, or fringed ones, as in Soldanella ; and if the process be earried to an extreme, we have ten petals in the place of five, as is not uncommon in Primroses and Cyclamen? (Pl. XXXIX. fig. 10). This lateral branching of individual petals must not be confounded with the dialysis or complete separation of one petal from another, which has been observed in the flowers of Anagallis $ and other genera, and which is the normal condition in the genera Asterolinum and -Apochoris. Enation, or outgrowth from the surface of the petals, is a process scarcely distinguish- able from transverse chorisis except by the later period at which it occurs. Chorisis, as I understand it, is a congenital process, beginning with the growth and continuing * Quite as frequently, as far as I have observed in Lysimachia and Primula, the pistil originates as a ring, which grows upward from beneath so as to form a tube, upon whose edge the five tubercles are not formed till quite late in development, and in some cases, especially in monstrous flowers, not at all (Pl. XL. fig. 4). T * Vegetable Teratology,’ p. 359. | $ Г. с. p. 73. THE MORPHOLOGY OF THE РЕМОГАСЕХ. 289 with the development of the organ affected (unless arrested), whereas enation does not show itself for some time after the origination of the part. Chorisis is a true subdivision of the original organ. Enation is a secondary process, and is, as its name implies, an out- growth from an organ already advanced some stages in development. Such outgrowths are of frequent occurrence in Primroses ; and they have especial significance with reference to the mode of formation of the stamens, before alluded to (Pl. XXXIX. figs. 2 & 3), and also to the formation of the placenta. It is curious to observe that in some of these cases the dorsal face of the outgrowth is applied to the upper surface of the petal from which it springs, just as the leaves of a book follow one another in parallel order, as in the diagram, where the thick line represents the outer surface of the petals, and the thin lines the inner surface respectively =, while in other cases the In the latter case it upper surfaces are placed opposite one another, thus — — - ж, sometimes happens that the margins of the outgrowth adhere to those of the original petal, and a tube is formed. It must be understood that there is no real change of position in this latter case, such as could only be brought about by a twisting of the lower part of the outgrowth, but that the cellular structure characteristic of the outer surface is developed on the inner surface, and vice versd. A tubular or cupped condition of the petals is one of the commonest deviations in Primula prenitens t, the Chinese Primrose of gardens. It usually arises by a process of growth similar to that by which peltate leaves are formed; and as the marginal portions grow faster or more vigorously than the central ones, а сир or tube is eventually formed. In other cases the tube appears to be developed by the gradual ineurvation and ultimate cohesion of the margins. It is possible that this latter condition represents the formation of the anther-lobes. Adverting now to the principal malformations affecting the corolla of Primulacez as a whole, mention may be made of that general leafy condition known as phyllody or chlo- ranthy, and which is frequently accompanied by prolification 2. These changes will be incidentally referred to in the sequel. Another very common change consists in the dupli- cation of the corolla. А **hose-in-hose" corolla is thus formed $. In such cases the petals of the inner corolla generally alternate with those of the outer; and then their autonomy may be safely assumed: but in other cases they are superposed; and then they are probably the result of enation, as before explained. ° Аховастом.-—ТЬе commonest changes to which the stamens of Primulacez аге sub- jected are various degrees of leafy or petaloid mutation, phyllody or petalody, as the case may be, such as have just been referred to, and, more rarely, of detachment from the tube of the corolla ||. : es Enation.—When what should be а stamen assumes a leafy or a petaloid condition it very often becomes subject to chorisis or to enation, in the same manner as in the * Mast. in Journ. Roy. Hort. Soc. 1866, vol. i. new ser. p. 202 ; Veg. Teratology, рр. 377, 449. T Veg. Terat. p. 315. + See Marchand, Adansonia, 1864, p. 150; Baillon, Ann. Sc. Nat. sér. 3. t. 2, p. 290; Adansonia, iii. t. iv. ; een Bull. Soc. Bot. Fr. x. p. 191. $ Vegetable Teratology, p. 377. | Ibid. р. 253, 2 290 DR. М. Т. MASTERS ON SOME POINTS IN case of the petals just referred to. (Pl. XXXIX. figs. 19 to 23 show illustrations of this charaeter from a variety of Polyanthus known in gardens as the “ Nigger.") In some of these cases enation takes place from both surfaces of the leafy or petaloid stamen мфу (fig. 22), as may be illustrated thus | , the central line indicating the axis of the staminode, the lateral ones the position of the outgrowth. In some cases the flowers are still further complicated by the development of these supplementary outgrowths in the form of horns and tubes, either by cohesion of the margins, or by unequal growth and consequent depression of the centre, as before explained. Pistillody of the Stamens.—Much the most interesting change, however, is that in which the stamens assume more or less of the condition of the carpels. The first case of the kind that fell under my own observation was nearly a quarter of а century ago, when I had the opportunity of inspecting a drawing and description of some flowers made by the Hev. G. E. Smith. His figures and his descriptions were so remarkable, that in my * Vegetable Teratology’ I confined myself to a mere mention of the case, and waited the time when, by the inspection of similar flowers, I might be able to form my own interpretation of the appearances presented. Mr. Smith's drawings bore every mark of accuracy; but I might well hesitate as to the correctness of the explanation given, when I read the following descriptions of the drawing :—‘“ Stamen with anther sitting upon an imperfect petal and bearing a style; enlarged pollen with terminal style; anther bearing one long twisted style and stigma; stamen with pollen-cell pro- liferous, &c.” This was а description to whet the ardour of the humblest morphologist; but it was not till a short time ago, through the kindness of my friend Mr. Alfred Bennett, that I received from Miss Dowson, of Beccles, specimens evidently closely similar to those of the Rev. G. E. Smith, sufficiently so to vindicate the correctness of his drawings, and to supply an explanation of what was paradoxical іп his descriptions. А comparison of the figures (Pl. XXXIX. figs. 24—82, which are copies from the drawings of the Rev. G. E. Smith), together with others taken from Miss Dowson's specimens by Mr. Worthing- ton Smith and by myself, will show how close is the correspondence between the two sets of specimens (Pl. XXXIX. figs. 34-39, Pl. XL. figs. 1-3, and Pl. XLI. figs. 1-8). No two of these flowers were exactly alike in all particulars; but there was so much general similarity between them in essential points, that 16 is not necessary to enter into details of particular flowers. Тһе general appearances presented may therefore be _ briefly summarized as follows :— _ Calyx normal. Corolla of a variable number of free, or nearly free, linear-oblong petals (Pl. XLI. figs. 1-3), the increase in number being obviously due to the fission or lateral ehorisis of the original petals. This was seen by the fact of there being in some of the flowers, where the liberation of the petals was not complete, a short tube with five hor eese each speedily dividing into two for the supply of the lateral sub- divisions ' THE MORPHOLOGY ОҒ THE PRIMULACEA. 291 In the place of the normal stamens, but wholly detached from the corolla, were numerous free hypogynous pistilodie stamens, showing various degrees of transition between the conformation of the stamen and that of the carpel (Pl. XLI. fips. 4-7). The filamentary portion of these pistillodie stamens was, in some cases, but little changed ; in other instances it was more or less dilated and leafy, and often crumpled or spirally twisted. In some the filament was dilated into an open carpel, destitute of ovules, but prolonged into a long style terminated by a stigma (РІ. XL. fig. 2). The ovarian portion was sometimes absent; and then the appearance was that of a style only (Pl. XL. fig. 3). In other eases the midrib of the filament or its central vaseular bundle was thickened, and bore imperfect ovules (Pl. XX XIX. figs. 33-39). It is especially noteworthy that this placenta-like rib, while generally congenitally adherent to the leafy or carpellodic filament, was not unfrequently detached from it by chorisis, the significance of which fact will be further insisted on in the sequel. In another series of these flowers the dilated filament was irregularly and pinnately lobed, the terminal lobe being incurved—a general occurrence, indeed, in the flowers in question. Тһе uppermost of the lateral lobes very often bore marginal ovules, while the lower ones were infolded at the edges and prolonged into a long style and stigma (Pl. XL. fig. 1). Here, then, we have a branched or divided carpellary leaf with some of its lobes open and others closed, both ovule-bearing. We have, in fact, the exact homologue of the so-called compound or branched stamen. Тһе bearing of these facts on the nature of the free central placenta will be made apparent hereafter. In the illustrations just cited it was the filamentary portion which was especially involved; but in another set of flowers the filament was unaffected or nearly so, and the changes were confined to the anther. In the Rev. G. E. Smith's sketch the filaments were all normal, while in Miss Dowson's specimens they were usually more or less dilated. Тһе changes observed in the antheral portion of the stamen, if stamen it can be called, were similar to those described in the case of the filament. Of anther, properly speaking, there was little or no trace, though ‚ the central placentary rib may be taken as the homologue of the connective, the latter prolonged occasionally into a style. In no case was I able to find any pollen at all. Тһе ovules, indeed, occupy the situation of the pollen; but it could not be expected that there should be any transition between the pollen-cells and the оущез. Nor was I able to find, as I had hoped I might do, any trace of pollen in the ovules, as I have before seen in Passiflora and Rosa. It should further be noted that the pistil in these flowers was quite normal *. * Singular as these bisexual phyllomes are, they are relatively by no means unfrequent in certain plants, such as Sempervivum tectorum, Cheiranthus cheiri, and other plants mentioned in teratological records (see ‘ Vegetable Tera- tology,’ p. 319). Even Conifers occasionally show similar phenomena. Thus, at the time of writing this paper, Mr. Syme, of the Elvaston Nurseries, sent me a catkin of Cupressus Lawsoniana in which the lower bracts were antheriferous and the upper ovuliferous; and, more than this, one of these bracts (РІ. XLI. figs. 9-11) bore on the outer surface, at the lower edge, an anther-cell, while on the inner face was a shortly stipitate ovule. (Ersted has recorded similar phenomena. 292 DR. M. T. MASTERS ON SOME POINTS ІМ ух жегом.— P yllody, or leafy condition of the pistil, is of not infrequent occurrence in the Primrose. Imperfect development is also not uncommon, in which case an imperfect tube or sheath, open at the top without trace of style, surrounds the placenta, the latter being in such cases usually elevated on a long stalk (Pl. XL. fig. 4). "This is in accordance with the natural development of the carpel, save that it is arrested. In other cases, especially in the Chinese Primrose (P. prenitens), the carpels assume а petaloid guise, the entire pistil being completely open at the summit (Pl. XL. figs. 8 & 9). Dialysis, or a lack of union of the carpels, is of frequent occurrence in the Primrose, there being in the simplest cases five open carpels, each with its separate style and stigma. Ina double Polyanthus lately shown at the Royal Horticultural Society, the calyx was normal, the corolla in two rows, the first row normal, with stamens as usual. Within them was a second corolla, also provided with stamens. Тһе gynscium con- sisted of five more or less detached petaloid open carpels, each with a long style and stigma. One of these carpels was remarkable for having its margins divided into lobes; of these lobes the uppermost were prolonged into styles (Pl. ХТ, fig. 11), the lower bore imperfect ovules; so that the same сатреПагу leaf bore secondary carpels and ovules, exactly as in the case of the pistillodie stamens previously mentioned (Pl. XL. fig. 1). PLACENTATION.—The free central placenta of Primulacez has naturally attracted much | attention on the part of morphologists, some of whom have considered it to be of axial, some of foliar nature. The former view is that which is most generally adopted, though of late it has been contested by men of authority, on which account it may be well briefly to allude to the most important papers that have been published on the subject. A. de Saint-Hilaire * describes the placentas as originally connected with the interior of the style by a little thread, which is afterwards obliterated. Duchartre, in the paper before referred to, combats this view, and says that in the development of the ovary and the placenta “ Гоп remarque deux développements marchant parallélement et simul- tanément; celui de la portion appendiculaire ou des parois de l'ovaire, du style et du stigmate, et celui de la portion centrale ou axile. Celle-ci pendant toute la durée de son développement reste libre et indépendante de la partie externe du verticille femelle; elle joue absolument le róle d'un petit rameau végétant sous un abri protecteur, le seul rôle du reste qu'elle puisse jouer. D'abord entièrement homogène elle se laisse diviser plus tard en deux parties, l'une inférieure, qui donne naissance aux ovules et qui possede toute la structure de l'axe lui-méme, l'autre supérieure et stérile uniquement celluleuse, dont le développement est, le plus souvent, trés-horné, qui parfois s’accroit . . . et devient un petit cône logi dans la partie inférieure du canal stylaire."' Payert entirely adopts the conclusions of Duchartre, as, "m do most modern observers. _ А large number of teratological phenomena might be cited in favour of this view, such as the instance cited by Duchartre in which a flower-bud was observed at the apex of the placenta i in Cortusa Mathioli. Other prolifications of a similar character have been p _ Mémoire sur les plantes auxquelles on attribue un placenta central libre (1816). à Е inogé nie Végétale, p. 611. THE MORPHOLOGY ОЕ THE РЕТМОТАСЕХ. 293 recorded by Marchand and others. Baillon * even mentions a species of Lysimachia in which the placenta was prolonged beyond the ovary in the form of a shoot bearing leaves, and which was cut off and struck as a cutting. Alphonse de Candolle and many others have noticed the placenta giving off leaves instead of ovules. : In double Primroses it is not unusual to find the placenta bearing ovules at its lower part and one or more carpels above (Pl. XL. figs. 12-15)—a condition which has its parallel in the singular instance described by Mr. Berkeley in a species of Dianthus +. Such instances as those just alluded to support the view that the placenta is an axial structure; nor is that notion necessarily invalidated by the very common occurrence of free central, marginal, and parietal placentation, with every intermediate state, in the ovaries of double Primroses; for the presence of ovules, as of buds, is of itself no proof either of the foliar or of the axial structure of the organ from which they spring. There are, however, many valid reasons for supposiug the placenta to be of foliar origin : thus Van Tieghem 1, who at one time considered the placenta of Primulacez to be the direct prolongation of the axis (Ann. Sc. Nat. 5* sér. 1868, t. ix. p. 211), is now satisfied that the vascular bundles of the placenta have their spiral vessels on the outer side, and that they are therefore appendicular. Оп the other hand, he admits that in _ cases of virescence or prolification, the axial system of the pedicel is abnormally pro- longed into the cavity of the carpels. In this latter case the trachee are internal. The vascular bundles of the placenta then, instead of being independent of the carpels, are offshoots from them. А similar conclusion is arrived at with reference to Theophrasia. Ina note appended to p. 622 of his translation of Sachs's * Lehrbuch,’ M. Van Tieghem gives the following summary of his present views, which it is convenient to cite in this place. * L'étude anatomique démontre que le pétale et l'étamine superposés (Primulaceze), пе sont que les deux parties d'une seule еб méme feuille ramifiée, qui а subi une méta- morphose héterogéne et double appropriée à deux fonctions différentes. Mais, en outre, elle fait voir que la colonne placentaire est formée, non par le prolongement de l'axe floral au dessus de l'insertion des carpelles, comme on l'admet généralement, mais par des dépendances liguliformes des carpelles qui constituent la paroi de lovaire unies et soudées entre elles dans l'axe géométrique de la fleur comme les сагреПев le sont eux- mémes latéralement dans toute leur étendue. De sorte que les ovules sont ici, aussi, insérés en définitive sur la feuille carpellaire, et qu'ils ont par conséquent la valeur morphologique de lobes de feuille "$. AL Braun || held somewhat similar views to Van Tieghem’s, and pointed out that many сазез in which the placenta appears to be axial may be explained by the congenital detachment of the bases of the carpellary margins and their fusion one with another ж Baillon, * Adansonia,’ iii. p. 310. + Gard. Chron. Sept. 1850, p. 612, and Veg. Terat. р. 268. + Ann. Se. Nat. 5° sér. t. xii. p. 329 (1871); ‘ Recherches sur la structure du pistil’ etc. p. 91. $ Tréeul, however (Comptes Rendus, Oct. 23, 1876), contests Van Tieghem's views, and shows that in Anagullis the vessels of the placenta originate independently of and before those of the walls of the ovary. The parietal and placental vessels do not correspond in number ; nor are they continuous with, but quite free from, those of the thalamus. || Analysis in Bull. Soc. Bot. France, Rev. Bibl. (1875) t. xxii. p. 63. SECOND SERIES.—BOTANY, VOL. 1. 2T 294 DR. M. T. MASTERS ОМ SOME POINTS ІМ in a line continuous with the floral axis, although he admitted that in Primulacez there are grounds for supposing the placenta to be truly axial *. 'The appearances presented by the double Primrose sent by Miss Dowson substantiate the view of Van Tieghem, that the placenta is derived from the carpellary leaf by chorisis (Pl. XXXIX. figs. 34-38). In the specimens in question the placentas are sometimes marginal, at other times parietal, the placenta in this latter сазе being homologous with the midrib of а carpellary leaf, and not unfrequently detached from it, sometimes slightly, at other times completely (Pl. XXXIX. figs. 37, 38). Double- flowered Primulacez in fact frequently furnish illustrations of free central, axile, mar- ginal, and parietal placentation (Pl. XL. fig. 16). Moreover they show that in some cases the placenta, whatever its position, is a mere outgrowth from the carpellary leaf, capable of becoming completely detached from it, and in that case simulating an axis. In a malformed Cowslip, a figure of which is given (Pl. XL. figs. 17-20), among other changes the pistil was broken up into its five carpels, each with its style and stigma. In the centre of the flower was а mass of threads or funicles, each surmounted by an imperfect ovule. Some of these threads evidently sprung from the inner surface of the сатре]з, others from the margins, while the majority were quite detached from the car- pellary leaf. In the normal flower this detachment must take place, if at all, at а very early period, and the placentary ribs of one carpel may become united to those of another, and so, at length, a free central placenta may be formed. If the axial appearance of the placenta and the five vascular bundles visible in it be cited as proof of its axial nature, it may be pointed out that five petioles, each with a single vascular bundle and united together, would give an appearance of an axis. For my own part, I look on the alleged distinction of axial and appendictlar organs as one of convention, desirable to be kept up for descriptive or classificatory purposes, but altogether of secondary importance from a morphological point of view. The occurrence of leaf-buds or flower-buds on the placenta, such as have been recorded in various monstrous Primulacee, therefore has not for me the same degree of significance that it has for those who look upon their presence as indicative of the axial nature of the organs bearing them. Leaf-buds, flower-buds, and roots may and do arise from leaf- organs or phyllomes, as they do from caulomes, and they cannot be looked on as abso- lutely characteristic of either +. * In the female flower of an Alder (Alnus) sent to me by Mr. Stratton, among other changes, the ovules were borne - on a stalk standing up free in the centre of a one-celled pistil, forming a free central placenta. It seemed probable that in this case the stalk was composed of the detached dissepiments of the ovaries. + AsI write Ihave before me a flower of Calochortus sent me by Mr. Burbidge, in which the filament of the н bears at its junction with the anther a well-marked bud (Pl. XL. fig. 21). These adventitious buds are sometimes veritable enations; at other times they are borne on stalks which are congenitally adherent to to the leaf- organ. Зее Planchon, * Quelques mots sur les inflorescences epiphylles,” Mém. Acad. Stanisl. 1862, p. 403; Caspary, Schrift, phys.-ckon. Gesellsch. zu Kénigsberg, 1874. Still more germane to the subject now under consideration are those cases in which the placenta bears leafy ovules, as happens occasionally in Crucifers. In a species of Triumfetta, and in one of Corchorus, I have seen a similar arrangement, the latter case (РІ. XL. figs. 21-24) being | particularly ерт ев the position of the leafy ovuleson the midribs of the — not on the margins. THE MORPHOLOGY OF THE РБІМІЛ,АСЕ Ж, 295 So far, then, as the nature of the placenta is concerned, we have two theories : according | to the one it is essentially axial; according to the other it is foliar. Organogenetie evidence is strongly in favour of the axial view, though it is quite clear that, had it been practieable to observe the mode of development of the monstrous Primroses above described, the foliar nature of their placenta would have been proportionally as obvious as in the adult state. The evidence derived from minute anatomy is conflicting; and it is doubtful whether so much importanee should be attached to the position of the vascular bundles and of their constituent elements as Van Tieghem is disposed to attribute to them. Comparative morphology is entirely in favour of the axial view ; there is not, so far as lam aware, a single genus of Primulacez or of Мутвіпасеге where, under normal cir- cumstanees, the placenta is any thing else but ** free central." Teratology, as we have seen, affords support to either view, and lands us in this dilemma. Either we must say that any or all forms of placentation, foliar and axial, may occur in Primulaceze—a conclusion to which morphologists would certainly demur, except as a monstrous condition; or we must admit that the evidence afforded by teratology is utterly worthless in deciding such cases, and that the appearances in question are, in fact, simply accidental and abnormal. То this latter conclusion equally strong objeetion may be taken. Extraordinary as these appearances sometimes are, they may generally be easily referred to ordinary morphological laws. If this be admitted, as it must surely be by any one who has paid any attention to the subject, then we are sent back to the morphological dilemma above mentioned. As a provisional means of dealing with it, it may be suggested, as an hypothesis, that the stock from which Primulacee descended had parietal placentation, but that in course of time the placentas have become detached from the walls of the ovary, and fused together to form the free central placenta. Looked at in the light of this hypo- thesis, the monstrous flowers above alluded to, showing every gradation between саг- pellary and axial placentation, are mere reversions to an ancestral form. For this hypothesis it must be admitted no evidence can be adduced beyond that of the monstrosities in question and the occasional presence of a little filamentous process (the filet of St.-Hilaire) connecting the placenta with the apex of the ovarian cavity—a process the existence of which is doubted by Duchartre, but which is nevertheless осса- sionally to be seen on making sections of young Primulaceous ovaries, and which is, on this hypothesis, to be explained as a rudiment of the process of chorisis, by which the former generations of Primroses changed their parietal for a free central placenta. OvuLEs.—From the ordinary appearance of the placenta in Primulacez, and the position of the ovules upon it, it might well be supposed that the ovules or, rather, their outer coats were independent of the carpels, and direct outgrowths from an axial pla- centa, each ovule being the representative of a distinct leaf or phyllome. The cases recorded by Unger, Alphonse deCandolle, Brongniart, Marchand, Cramer, and others *, * Vegetable Teratology, p. 263. 2T2 296 DR. M. T. MASTERS ОМ SOME POINTS ІМ may almost all be alleged in support of this view. Nevertheless, when the evidence thus presented us is analyzed a little carefully, and correlated with the facts revealed to us by anatomy and development, it will be found that in most cases the facts related are not incompatible with the notion that the placenta is essentially an outgrowth from the carpellary leaf, and, by consequence, that the ovules, or their outer coats, are also deri- vatives from that source. Although the question, whether the ovular coat be a direct formation from the axis, or whether it originates from the carpellary leaf, may still be said to be undecided, yet the foliar character of the ovular coat will hardly now be contested. To the establishment of this point the investigations of monstrous Prim- roses by Brongniart and others* have mainly tended. It is not necessary to do more than recall these observations, because their accuracy has never been called in question. The illustrations now brought forward show clearly that the ovules of these particular Primroses are processes from the carpellary leaf, either marginal or detached by chorisis from its centre. This is in accordance with the views of Van Tieghem as cited in the extract previously quoted. Тһе occurrence of secondary carpels on the placenta, mixed with the ovules, also confirms this view. Іп this latter case we have a branched or compound carpel, some of the lobes of which are developed as ovules, others as secondary carpels. "These remarks, of course, apply to the outer coat of the ovule, and not to the nucleus. None of the monstrous flowers of Primulacez that I have hitherto examined have given any clue to the nature of the nucleus. Summary.—Taking now a general review of Primulaceous structure, normal and abnormal, the facts cited seem to lead to the following conclusions :— 1. That the petals of most Primulaces (not in Samolus or Androsace) are late out- growths from the receptaeular tube outside the stamens, and posterior in development to them, but upraised with them by the upward growth of the so-called tube of the corolla. 2. 'That the placenta of existing Primroses is a direct prolongation of the receptacle or axis, without any connexion with the sides or apex of the carpels. 3. That the placenta in some (monstrous) flowers is an outgrowth, either from the margin or from the centre of the carpel, such outgrowth occasionally becoming wholly detached, the detached placentas, moreover, sometimes cohering one with another, and thus producing the appearance of a solid column directly prolonged from the receptacle. Such forms of placenta lead to the inference that the ancestral progenitors of Primulacec had parietal placentation, the monstrous forms being, on this view, looked on as reversions. _ 4 That both staminal and carpellary leaves may occasionally be divided or lobed, so that just аз we meet with compound or divided stamens in Malvacez and other orders, so the carpels of some orders are the homologues of divided leaves, as first pointed out ы Prof. Dickson. 5. That the ovular coat of Primulaces is essentially foliar, representing either the * Бе spocially Brongniart, Ann. бе. с dnb 308, and Ann. Sc. Nat. 3° sér. t. xi. p. 20 (1844); Cramer, Bildungsah еони THE MORPHOLOGY OF THE РВІМІЛ.АСЕ Ж. 297 blade (or dilated leaf-stalk) of an undivided ieaf, or, in other cases, a lobe or process » of the carpellary leaf, and not a direct production from the axis (assuming, for descrip- tive purposes only, a distinction between axis and appendage). 6. That the processes or lateral lobes of the carpellary leaf may be infolded, in some cases, so as to form secondary carpels. SELECT BIBLIOGRAPHY. The following are among the principal works and memoirs relating to the structure of Primulaceous flowers, and which have been consulted in the preparation of these notes :— Поснавтвк, “ Observations sur l'organogénie de la fleur &c. chez les plantes à placenta central libre,” Ann. Se. Nat. 3° sér. tome ii. (1844) pp. 279. Cortusa Mathioli, “Flower on apex of free central placenta.” ; Ввомамтавт, Ann. Sc. Nat. 2° sér. tome i. p. 308, t. ix., and 3° sér. tome ii. (1844) p. 20. Вавіметом, C. C., Ann. Nat. Hist. 1844, p. 464. Primula vulgaris. In place of the ovary а cup- shaped body, from the centre of which springs a cylindrical column, bearing a cup-like sheath, or im- perfect ovary at the top, surrounding a conical placenta studded with ovules. ALPHONSE DE CANDOLLE, Neue Denkschriften, Band v. p. 9. General phyllody of flower of Primula auricula, leafy condition of ovules, &c. P. prenitens, flowers within the ovary, &c. Baron ре Me icog, Ann. Sc. Nat. t. x. 1846. Prolification of inflorescence, &c. Uxszn, Act. Acad. Nat. Cur. 22, ii. (1850) p. 543, t. 5 & 6; Bot. Gazette, Мау 1851, p. 70. Pri- mula sinensis, phyllody of the parts of the flower, &c. Payer, Organog. р. 611, t. 153 (1857). Masters. Primula sinensis (March 1859). Pistil entirely wanting ; in place of it an umbel of about six rays, each bearing a small flower-bud. . Cramer, Bildungsabweichungen, р. 132 (1864). Masters, Vegetable Teratology, 1869. In this volume a large number of references are given to various malformations recorded by various observers or by the author. Cave, “Sur le placenta central libre des Primulacées” (analyzed in Bull. Soc. Bot. France, xvii. 1870, р. 110, Rev. Bibl.). The author supports the axial nature of the placenta on anatomical grounds. If, says M. Cave, the spiral vessels are on the inner side of the generating layer, the organ is axial, and vice versd. Еліуве, Sur Роуше et sa nature morphologique chez le Primula sinensis. Pamphlet, analyzed in the Bull. Soc. Bot. France, lxvii. (1870), Rev. Bibl. p.90. The ovule is considered to be homologous with a leaf. the funicules=petiole, the ovular coat=the lamina. See also Bull, Soc. Bot. Fr. 1869, p. 127. Cros, Essai de Tératologie Taxinomique (1871), p. 20. Іп this are cited the teratological records relating to the Primulacez, extracted from various publications issued since the date of Moguin-Tandon’s classic work, * Eléments de Tératologie Végétale’ in 1841. Prerrer, “ Zur Blüthenenttwickelung der Primulaceen,” Pringsheim’s Jahrbuch, viii, p. 194. Van Тчконем, “Structure du pistil des Primulacées," &c., Ann. Sc. Nat. 5* sér. tome xii. (1871) p. 339. J. L. pe Lanessan, “ Sur le développement des faisceaux fibro-vasculaires dans les organes floraux des Primula,” Bull. Soc. Linn. Par. No. 4, Nov. 1874. Shows, with Trécul, that the fibro-vas сшаг bundles are late formations in comparison with the cellular tubercles in which the several organs of the flower begin. 298 DR. M. T. MASTERS ON SOME POINTS ІМ Braun, Verhandl. des bot. Vereins der Provinz Brandenburg, 1874, p. 45. Сеглкоувку, Flora, 1874, p. 170; and Vergleichende Dartstellung der Placenten in den Fruchtknoten der Phanerogamen, 1875. Placenta of Primulacez considered to be axial. Ercuter, Blüthendiagramme (1875), p. 327. Considers that the outer row of stamens is abortive, the inner congenitally united with the petals. The placenta is considered axial, and the ovules foliar. Morgen, C., Bull. Acad. Roy. de Belgique, t. xix. No. 8, p. 539, c. tab.: Primula prenitens, vires- сепсе and phyllomorphy of flower. Bull. Acad. Roy. de Belgique, xix. p. 97, с. tab.: calycanthemy. DESCRIPTION OF THE PLATES. Puare XXXIX. Figs. 1-9 inclusive represent various Stages in the Development of the Flowers of Primula japonica. The figures are all highly magnified. Fig. 1. The young flower, seen from the side, with its bract b; c, the calyx; st, the staminal tubercles. Fig. 2. Flower, seen from the top. "The calyx is formed, and also the staminal tubercles, but no petals or pistils. Fig. 3. View of a fragment of a flower, showing three sepals and two stamens, with indications of the petals, p, at their base. Fig. 4. Side view of a flower rather further developed. The sepals are removed ; the staminal tubercles, ‘st, are shown, with the petals, p, and the commencement of the corolla-tube. Fig. 5. Portion of a flower still further developed, seen from within, and showing three staminal tubercles and the commencement of three carpellary tubercles, c. Fig. 6. Fragment of a flower in a further stage of development. Three carpels are shown free above, but congenitally coherent below ; the placenta, pl, is already formed in the shape of a dome-shaped mass, studded towards the base with ovules, which develop from above downwards. Fig. 7. Section of the young pistil of another flower, with the placenta, pl, not yet producing ovules, although the development of the pistil is further advanced than in the preceding. Fig. 8. Pistil in a further stage of development, and nearly closed. Fig. 9. Diagrammatic section of a flower with вера! (s), stamen (st), petal (p), carpel (c), and placenta (pl), all formed, and the upraising of the so-called corolla-tube just begun. Figs. 10-23. Illustrations of Chorisis and of Enation in the petals and stamens of Primula vulgaris etc. АП slightly enlarged. _ Fig. 10. Lateral chorisis of the petals; the venation at the base shows that each of the original vascular 1 bundles destined for the petals is split into two divisions. | Figs. 11-16. Enation; various outgrowths from the petals. 16. Section showing plan of arrangement | of parts in fig. 15. Fig. 17. Supernumerary petals in Primula Auricula. Fig.18. Sectional plan of the same. | б ET Stamen of Primula vulgaris, with a petal growing from it, and two supernumerary petaloid out- _ growths, developed in the order shown by the numerals (1. п. п. ту.) in fig. 20. Fig. a. Petal of Primula vulgaris, with petaloid anther. Fig. 21 а. Similar anther detached. Fig. 22. 2. Petaloid stamen of a variety of Polyanthus, “the Nigger,” developing from either side, by | ition, secondary petaloid laminz in the order in which they are numbered (1. to у. ). Fig. 23. Plan of he sa мец 21 the Roman numerals corresponding. THE MORPHOLOGY OF THE PRIMULACE 299 Figs. 24-32. Copied from drawings by the Rev. G. E. Smith of monstrous flowers of Primula vulgaris. Figs. 24, 25. Flowers, seen from the side: nat. size. Fig. 26. Flower, natural size; sepals and petals turned down to show the hypogynous pistillodic stamens and the normal pistil. Fig. 27. Normal pistil, with hypogynous petal, attached to which is a pistillodic stamen: magnified. Fig. 28. Pistillodic stamen from 26, magnified, showing a filament, a dilated anther bearing rudimentary ovules in the centre and at the margins and with a terminal style. Fig. 29. Pistillodic stamen with ovules, and three secondary carpels with styles. Fig. 30. Back view of the same, the styles cut across : magnified. Fig. 31. Filament with pistillodic anther, from the side: the margins bear ovules; and the apex termi- nates in a style. Fig. 32. Filament with pistillodic anther, with abortive ovules and terminal style. Figs. 33-38. Pistillodic Stamens from Primula vulgaris, taken from the flowers shown in Plate XLI. : all slightly magnified. Fig. 33. Pistillodic s stamen, with apex infolded, and with central placentary rib bearing ovules. In other cases this rib was quite detached. Fig. 34. Section of preceding. Fig. 35. Similar stamen to 33, with detached filamentous outgrowth. Fig. 36. Another specimen, placental rib quite detached except at the base. Fig 37. Section at upper portion. Fig. 38. Section at lower portion. PLATE XL. Figs. 1-3. Primula vulgaris, from Miss Dowson’s specimens (magnified). Fig. 1. Pistillodie stamen irregularly pinnately lobed, terminal lobe infolded, laterallobes some destitute of ovules, others ovuliferous on one or both margins; lowermost lobe infolded so as to become a secondary carpel, with ovules on its margin, and its apex prolonged into a style; from the centre emerges a placentary rib bearing an imperfect ovule. Fig. 2. Pistillodic stamen, in the form of an open carpel, with ovary, style, and EST stigma, but destitute of ovules. Fig. 3. Pistillodic stamen represented by a spiral style only. Figs. 4-16 show Leafy or Pelal-like Condition of Pistil etc. Fig. 4. Leafy pistil laid open, taken from a virescent Primrose, magnified. Тһе pistil, owing to arrest of development, was in the shape ofa tube open at the top. The placenta is seen above the orifice of the tube of the pistil raised on a long stalk, and covered by imperfect stalked ovules. Fig. 5. Section of upper part of placenta of the same. Figs. 6 and 7. Abortive ovules from the same. Fig. 8. Petalodic pistil of P. vulgaris open at the top, and cut down the side to show the placenta : magnified. Fig. 9. Similar pistil from P. prenitens. The free edge of the petaloid carpels is here fringed; the surface bears a few imperfect ovules (parietal placentation) in addition to the free central placenta. Fig. 10. Extremity of placenta from the same flower, enlarged, and with some of the ovules removed to show the apex of the placenta giving off thickened funiculi destitute of ovules. Fig. 11. Open carpel of Polyanthus, showing carpel lobed at the sides—the lobes being, some of them ovuliferous (ov), others prolonged into styles, as is also the apex : magnified. 300 ON SOME POINTS IN THE MORPHOLOGY OF THE PRIMULACE.. Fig. 12. Placenta bearing ovules (one of which is detached and enlarged at fig. 13), and a carpel with a long style and spatulate stigma. Fig. 14. Placenta with ovules removed to show the carpel. Fig. 15. Base of carpel. Fig. 16. Pistil of Primula prenitens, showing marginal and free central placentation, and dialysis of four styles: magnified. Figs. 17-20. Virescent Cowslips etc. Fig. 17. Flower of Cowslip, Primula veris. Petals virescent, stamens imperfectly developed, pistil dialyzed, ovules represented by thread-like funicles bearing abortive ovules. Fig. 18. Section of a similar flower. Fig. 19. Diagram showing the dialysis of the carpellary leaves and the position of the abortive ovules in figs. 17 and 18. Fig. 20. Section of normal ovary with free central placenta. Figs. 21, 22. Calochortus, sp. Figs. 21, 22. Front and back views of a petaloid stamen of Calorchortus, sp. Тһе anther is petaloid ; and . at the junction with the filament is a bud, three leaves of which are petaloid, unfolded, and spreading ; the others, smaller, are still undeveloped. (Drawn by Mr. Burbidge.) Figs. 23, 24. Corchorus, sp. Fig. 23. Flower of Corchorus with five sepals, the petals and stamens removed to show the stalked ovary. Fig. 24. Ovary laid open, to show leafy ovules on dorsal and ventral sutures of carpels. Рглте XLI. Figs. 1-8. Primula vulgaris with pistillodic stamens (Miss Dowson), drawn by W. G. Smith, Esq., F.L.S. ; АП magnified. Fig. 1 shows a longitudinal section, with numerous hypogynous pistillodic stamens, the filaments leafy or petaloid, the anthers ovuliferous in the middle, and the connectives prolonged into a two- lobed ineurved petaloid lobe. Тһе central pistil is normal, with a long style. -Figs. 2-3. Sections of similar flowers. Тһе apices and sometimes the lateral lobes of the pistillodic stamens are prolonged into styles terminated by stigmas. Fig. 4. Pistillodic stamen from one of the above flowers. Filament dilated; anther open, ovuliferous in the centre; connective petaloid, two-lobed. Figs. 5—7. Similar stamens to fig. 4. Fig. 8. Pistillodic stamen reduced to a style with terminal spatulate stigma. Figs. 9-11. Cupressus Lawsoniana. Fig. 9. Back or outer surface of bract with anther-lobe at lower edge. Fig. 10. Front of the same bract, with a single ovule raised оп a short stipes. Тһе usual ovuliferous scale is abortive, so that the ovule appears to arise directly from the base of the bract. Fig. 11. Longitudinal section through the bract, to show the position of the anther and of the оуше: a, anther; бу, bract; о, ovule. L Trans Linn. Soc. SER 2. Вот M1 Masters. >. ы (J ә < на c: 2 ac [om Б- СУ z M < = td e |= 27 e С © B TRANS. LINN. бос SER 2Вот Мог 1 ТАВ 40 MT.Masters. Е z i Е 3 - о [x 2, о о tá О +2 E. JNFitch lith. TRAms.Liwm.Soc SEeR.2.Bor VoL I Тав 4]. M.T. Masters. - Fitch пор. FLORAL CONFORMATION OF PRIMULACE Æ, ег. JNFitch Б.