Imágenes de páginas






(PLATES 330, 331). The characters employed in grouping the different species of the genus Cycas are not altogether satisfactory. No doubt this is due to the absence of complete materials for the knowledge of the species, either in cultivation or in herbaria. The portions of the plants necessarily best known are the leaves; they have been employed as a basis of classification, the characters depended upon being the revolute margins, or the more or less flat nature of the segments. But the fact that in the most characteristic revolute species (C. revoluta L.) there are plants with flat margins shows that this can be of little value, while the revolute species from India (C. Beddomei Dyer) has its affinities with the C. circinalis and the Australian species, and not with C. revoluta or C. inermis. Neither can the

presence of tomentum on the spadices be of much value, as this depends in several species on the age of the spadix. It appears to me, looking at the materials existing in the Herbarium of the British Museum, and at the published figures and descriptions, that the form of the barren expansion in the female spadices will supply, in the present state of our knowledge, better characters for grouping than any hitherto suggested. Three types are present :

First. Where the apex is dilated into a rhomboidal lamina, with teeth on the two upper margins of the rhomb, the terminal one being usually much larger. To this group belong C. circinalis, C. Pumphii, C. Seemanni, the Australian species, &c.

Second. Where the lamina is longer than broad, and is deeply cut along the sides into spiny teeth. To this belong C, revoluta Linn., C. inermis Lour., and C. Taiwaniana here described.

Third. Where the lamina is broader than long, and the spiny teeth are borne chiefly on the upper margin.

To this group belong the species discovered and figured by W. Griffith,-. pec. tinata Griff., C.Jenkinsiana Griff., C. macrocarpa Griff. The materials for the history of this group are still very imperfect.

In the herbarium of Dr. Hance, which was some years ago JOURNAL OF BOTANY.--Vol. 31. [JANUARY, 1893.]


acquired by the British Museum, there is part of a leaf and three foliar spadices of a Cycas from the Island of Formosa. It belongs to the group of C. reroluta, though the barren lamina approaches the species of the first group. It may be thus described :

Cycas Taiwaniana, sp. nov.-Leaf with numerous erectopatent subopposite segments, springing from a terete rachis; petiole unknown ; segments flat, linear-lanceolute (5 to 7 in. long, rather more than 1 in. broad), decreasing below to a base about half the width of the seginent, decurrent, but scarcely turned upwards on the rachis, shining, paler on the under surface.

Male cone unknown. Female spadices nearly glabrous, long, with slender stipes; fruit (3 or 4) borne above the middle; lamina nearly as broad as long, deeply cut on both sides into linear acuminate spines of the same substance as the lamina; terminal spine somewhat longer, broad and serrate.

The specific name is from Tai-wan, the native name of Formosa. No more definite information is contained on the label than that the specimens were collected in the island of Formosa by Mr. Swinhoe, and sent to Dr. Hance in the autumn of 1867, from whose herbarium, as I have said, came the specimen in the British Museum on which the species is founded.

In the Flora Vitiensis Dr. Seemann described a Cycas which he found in the Fiji Islands, and referred to ('. circinalis L. A. Braun subsequently pointed out characters by which he separated it from C. circinalis L. and named it C. Seemanni. Baron von Mueller has described the plant at length. Dr. Masters having lately given the Botanical Department a series of photographs of the plant, it seemed to the Editor desirable to give an illustration of this fine Cycad, discovered by and named after the founder of this Journal. It has a stem thirty feet high. In the specimen figured from the photograph, an adventitious bud, developed two-thirds up the stem, has maintained its connection with the stem and developed into a branch. The stem is marked by alternate constrictions and enlargements, caused by the alternation of the fruiting spadices and the normal leaves. The scars left by the spadices are smaller, and these being food-consuming organs, the stem is constricted where they have been borne. The petiole is unarmed, and the numerous segments (50 to 70) are papyraceous, spreading and curved; they gradually decrease from a little above the constricted base, and end in a long acuminate apex. The male cone is two feet long, and the scales have a short, acute, ascending apex on the upper part of the cone. The female spadix bears from six to eight fruits ; it has a dilated, subtriangular apex, with small spines along the upper margins and a terminal one scarcely larger than the others. It was found in Viti-Levu and Ovalau by Dr. Seemann.

In the Museuin Herbarium there are specimens of a Cycad from the Tonga Islands, collected by Banks and Solander in Capt. Cook's first voyage, which was referred by Dr. Seemann with his Fijian plant to c. circinalis L. It differs in the texture and form of the segments of the leaves, and the presence of a large terminal spine on the spadix; but until more materials are obtained from the Tonga,

Fiji, New Caledonia and neighbouring islands, it is undesirable to add new names to the genus, as they may represent only unimportant geographical modifications.

EXPLANATIONS OF PLATES. Tab. 330.-Cycas Seemanni A. Br. Representing the general aspect of the plant, the male and the female fruiting heads, with a single spadix, all somewhat reduced in size from photographs.

Tab. 331.-Cycas Taiwaniana, from specimen in the British Museum.




(Continued from vol. xxx., p. 311.) GROUP 8. BELLARDIANI (= GLANDULOSI Focke).--St. mostly prostrate and roundish, rooting, often glaucous. All the stems densely clothed with stalked glands, bristles, acicles and prickles of various sizes. Prickles more frequently weak than in the RadULÆ and the KOEHLERIANI ; often subulate. Pan. racemose or with racemose lateral branches at the base. All the lts, distinctly stalked. Stipules filiform. Stam, rather frequently falling short of the styles, or barely equalling them.

Usually rather small low-growing plants.

A. Stalked glands very unequal; some of those on the pan. longer than the diameter of the ped. :-(74) viridis ; (75) Durotriyum ; (76) divexiramus ; (77) saxicolus ; (78) Bellardi ; (79) serpens ; (80) hirtus and vars. All nearly allied plants.

B. Stalked glands short; those on the pan. hidden in the dense hair or felt (" sunken "), or at least shorter than the diameter of the ped. :--(81) tereticaulis ; (82) ? oigoclados and vars.

74. R. VIRIDIS Kalt. Journ. Bot. 1890, pp. 134, 166. R. incultus Wirtg. Syn. R. G. p. 369.- St., petioles, pan.-rachides and ped. all thickly clothed with very unequal prickles, acicles, bristles and stalked glands, usually densely hairy, more rarely almost glabrous. St. long, prostrate, roundish or bluntly angular. Prickles mostly short and declining, conical, broad based. L. cluietly 5-nate-pedate. Lts. pale green, shining above, more or less hairy on both sides, rather coarsely and irregularly but not deeply dentate-serrate; term. roundish or broadly ovate-cuspidate or elliptic-acuminate from a slightly emarginate base, often with 1 or 2 lobate dentitions above the middle (usually on one side only). Pari. usually rather long and lax, pyramidal, with straight rachis and numerous nearly patent few-flowered branches, clothed like the st. except in having still slenderer aciculate prickles. Sep. attenuateacuminate, purple with stalked glands, patent in fr. Pet, very long and narrow, pointed and cuneate-based, white or slightly pinkish. Stam. white (or reddening later), usually far surpassing the styles. In several counties (N. & S.).

When growing in woods, very similar to R. pallidus W. & N., but readily distinguished from it by its more unequal prickles, acicles and stalked glands and less diffusely branched pan., and also usually by its rounder, less acuminate, less deeply toothed and less cordate-based term. It. In open sunny places the plant be les much stouter, its 1. lose their soft hairs, and its broadly pyramidal and nearly naked panicles are enormous.

It then recalls the next species and rosaceus.

75. R. DUROTRIGUM R. P. Murray, Journ. Bot. 1892, p. 15.-St. prostrate, bluntly angular, apparently quite glabrous, yellowish on the under side, bright red above, densely clothed with slender acicles, bristles and stalked glands of all sizes. Prickles also remarkably crowded, very long-based, very slender, declining, falcate and deflexed. L. 5-nate-pedate to 3-nate, subpersistent. Lts. green, subylabrous, acutely doubly incise-serrate, acuminate; term. broadly roundishovate or slightly obovate, with long, gradually acuminate point and subcordate base. Pan. lax, with flexuose bairy rachis (armed like the st.) and crowded ultra-auillary rounded top; its lower 1. 5-nate. Dors.

R. Durotrigum seems nearly allied to the open-ground states of R. viridis (not yet found in Dors.), but it differs from them by its slenderer and far more crowded and still more unequal prickles and acicles and various gland-tipped organs,--the prickles also being longer-based and more variable in direction ; by the far less hairy Its., with their longer points and deeply incise-serrate acute toothing; by the more interrupted pan. and flexuose and still more strongly armed rachis; and by the small pinkish pet. Its sep. are attenuate and patent as in viridis, or somewhat loosely reflexed in fr., but they seem more hirsute, and so perhaps rather less conspicuously glandular. Its stam. are usually shorter and its styles proportionately longer, though apparently somewhat variable in length in both species. Its ultra-axillary branches are also usually more crowded together, and its lower branches more distant from them and from each other, so (together with the flexuose rachis) giving the pan. as a whole a less markedly pyramidal outline than in viridis.

So far found only in Dors., though in at least three or four distinct localities and in considerable quantity, and showing no noticeable variation under changed conditions of shade and soil.

76. R. DIVEXIRAMUS P. J. Muell. St., petioles, pan.-rachides and ped. armed and clothed much as in R. viridis. St. long, prostrate, slender, roundish, with rather many scattered and clustered hairs, dark purple. L. 3-5-nate-pedate, mostly 3-nate. Lts. yellowish-green, thin and flexible, subglabrous above, soft with many short shining hairs beneath, with acute crowded teeth, which are nearly simple in the 3-nate 1. but become more compound in the 5-nate; term. obovate-acuminate or cuspidate-acuminate, with narrowed and somewhat obtusangular truncate base. Pan. only slightly narrowing above into the conspicuously cylindrical ultra-axillary top, with many longish patent or even divaricate 1-3-flowered branches and subsessile term. fl.; the slightly flexuose rachis and the ped. more or less felted above, densely hairy, with many very slenuler aciculate prickles and crowded unequal acicles and stalked glands. Sep. triangular-ovate with very long points, externally green, aciculate and glandular, clasping fr. Pet. small (scarcely exceeding stam.), oblong, distant, white. Stam. white, exceeding greenish styles. Woods and bushy places (Glost., Heref., Dev.).

A distinct-looking plant; when fresh appearing just intermediate between R. longithyrsiger and R. viridis, and frequently growing with the former, though not observed by me with the latter.

77. R. SAXICOLUS P. J. Muell.--"St. angular, nearly glabrous. L. mostly 5-nate. Lts, with short soft hairs beneath, shining, especially on the nerves; term. broadly ovate, pointed. Inflorescence often elongated, lax; branches often with aggregated ped., densely patent-hairy, furnished with crowded glands, bristles and acicles. Sep. patent in fr. Pet. narrow, white." The foregoing is a translation of Dr. Focke's recently published description of this species. Speaking of its distribution in Germany, he adds, " The typical form is rare; but similar forms, approaching R. viridis, hirtus or Koehleri, are very common.” Plants from Oxf., Suss. and Monm. that he has thus named for me have brownish polished st., with very unequal broad-based prickles and acicles and comparatively few stalked glands, l. greyish green beneath, remarkably hairy pan.-rachis with most of the unequal-stalked glanuls hidden in the hair, the pan. branches crowded above into a rather narrow, rounded, cylindrical top, with short, distant, few-flowered branches below, and very small pinkish pet.

There is so much difference of opinion amongst us in England as to the distinctive characters of the three next “species," that it seems desirable for me in their case to give a translation of Dr. Focke's descriptions.

78. R. BELLARDI W. & N.?, R. dentatus Blox. “(R. glandulosus and R. hybridus autor. mult.).--St. only indistinctly angled near the top, glaucous, sparsely hairy, densely clad with unequal weak prickles, glandular bristles and stalked glands. L. 3-nute.

Lts. almost equal in size, light green, rather evenly and finely serrate, green and hairy on both sides; term. elliptic, with a lanceolate or linearlanceolate mucronate point. Inflorescence short; the lower branchlets erect-patent, usually 3-flowered; the upper straggling, 1. flowered; rachides and ped. hairy, with fine acicles, red with numerous unequal-stalked glands and glandular bristles. Sep. embracing the young fr. after flowering. Pet. narrow, spathulate, white. Stam. fully as high as the styles. Drupelets glabrous. Fr. small, aromatic." “In very few brambles,” Dr. Focke adds, “is the form so constant as in R. Bellardii ” (the spelling preferred by him);

hence it can readily be recognized everywhere, although the characters otherwise afford no distinct means of differentiation from the forms of the R. hirtus group. On cool, wooded soils, especially in springy ground.”

Prof. Babington's fuller description in Brit. Rubi, down to the middle of p. 248, agrees admirably with this; as both do with Welsh specimens of mine, which Dr. Focke refers here as “quite

« AnteriorContinuar »