fyxntll ^nivmii^ Jit«g PURCHASED FROM FUNDS OF THE FEDERAL EXPERIMENT STATION ^,.ur^./^ RETURN TO ALBERT R. MANN LIBRARY ITHACA, N. Y. Cornell University Library QK 495.A68S45 Revision of the natural order Hederaceae 3 1924 001 693 088 Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924001693088 REVISION OF THE, NATURAL ORDER HEDERACE^, BEIN& A EEPEINT, WITH NUMEROUS ADDITIONS AND CORRECTIONS, OF A SERIES OF PAPERS PUBLISHED IN THE 'JOURNAL OF BOTANY," BRITISH AND FOREIGN. BY BERTHOLD SEEMANN, Ph.D., F.L.S., PEES. ADJ. OF THE IMPEBIAL OEHMAN I. C. ACADEMY NATTIRiE OtIBIOSOEUM. Wiii^ Sllustrattoits. LONDON: L. REEVE AND CO., 5, HENEIETTA STREET, COVENT GAEDEN. 1868. CORN ELL- UN I VERSITYl FEINTED BT J. E. TAYLOR AND CO., LITTLB QDEXK STBE£T, LINC0LN''B INN 7IELDB. REVISION OF THE NATURAL ORDER HEDERACE^. From my ' Journal of Botany ' I here reprint, with numerous cor- rections and additions, a series of papers on all tJrabelliferoTis plants having valvate petals and a fruit composed of two or more carpels, of Which Hedera Helix is the. type, and for which I have adopted the name HederacecB; all genera with imbricate petals and 1-celled ovaries being excluded, as elements foreign to this group. LinnsBus has bequeathed to us only one genus of Hederacets {He- dera) ; his genera Panax and Aralia, reduced to their true limits, are not members of the group. In 1756, P. Browne published the genus SciadopJiyllum ; ten years later (1766), Forster, Schefflera and Polyacias. In 1780, Tbunberg established the genus Cussonia, and in 1789, Commerson that of Oastonia. In 1791, Geertner added Heptapleurum ; in 1802, Euiz and Pavon, Gilibertia and Actinophyllum, the latter identical with Scia- dophyllum; and in 1806, Petit-Thouars, Maralia. In 1830, De CandoUe enumerated thirteen genera, viz. Adoxa, Panax, Cussonia, Maralia, Oilibertia, Gastouia, Polyscias, Toricellia, Aralia, Sciadophyllum, Hedera, Paratropia, and Arthrophyllum. Five of these (Adoxa, Panax, Aralia, Toricellia, and Arthropliyllum') being excluded by the definition above given, the genera known to De CandoUe are reduced to eight, whilst the two published as new on this occasion were previously known. In 1840 Endlicher, in his ' Genera Plantarum' (including the first supplement), enumerated sixteen genera, adding to those known to De CandoUe Botryodendron, Miquelia, and Brassaia. But Miquelia belongs to Olacinea, and Botryodendron, previously described by Forster under the name of Meryta, as I have shown in 'Bonplandia ' for 1863, p. 294, to Haloraginece, an Order to which Helioingia, Willd., should also be referred. , In 1854 Asa Gray (Botany of Wilkes's Expedition) added PLeynold- sia, Tetraplasandra, and Plerandra, — all three excellent genera. In the same year (1854) Decaisne and Planohon commenced, in the 2 REVISION OF THE NATURAL ORDER HEDERACEiE. ' Eevue Horticole,' a revision of the Order " Araliacea" which unfor- tunately has never been completed, only nineteen genera being treated upon. It is a sketch, vrritten merely for a gardening paper, but ne- vertheless moat important. The learned authors there availed them- selves of several characters pointed out in 1848 by Koper, viz. the ar- ticulation of the pedicels, the calyculus, and the aestivation, in circum- scribing the genera. The new genera indicated were found, in most instances, to be so natural, and to coincide so well with the geographi- cal distribution of the species, that most of them have been generally adopted, though they have, as yet, not been described by them. They were eight in number (viz. Stilbocarpa, Echinopanax, Fatsia, Brassaiop- sis, Dendropanax, Oreopanax, Didymopanax, and Cuphoca^pus), two of which, Cuphocarpus, with its unicellular ovary, and Stilbocarpa, on account of its quincuncial petals, must be excluded from the Order, whilst EcJtinopanax has proved identical with Horsfieldia. In 1856, Miquel published in my 'Bonplandia' a paper on the Araliacea of the Indian Archipelago, in which he establishes five new genera (viz. Agalma, Eupteron, Aralidium, Macropanax, and Notho- panaio) ; and in a subsequent, though antedated publication, his ' Flora of Dutch India,' and its supplement, he adds to them Actinomoj'phe and Parapanax. Of these Aralidium, with its 1-celled ovary, will have to be transfewed to Cornacece ; whilst Eupteron is identical with the older genus Polyscias. In 1863, the same author published the genera Boiryopanaoc, Osmoxylon, Textoria, and Kalopanax, one of which only (Osmoxylon) proved new. In 1856, Hooker fil. and Thomson established the genus Tupidanthus ; in 1858, Grisebach ('Bonplandia') the genus Sciadodendron ; and in 1859, Maximowicz, the genus Eleu- therococcua. In 1859, Carl Koch gave, in his ' Wochenschrift fiir Gartnerei und Pflanzenkuiide,' an enumeration of the garden Araliacece known to him, and a review of the labours of those gone before him in this field of inquiry. Though overlooking a few genera, he enumerates thirty-four, of which seven (Aralidium, ArtJirophyllum, Cuphocarpus, Panax, Aralia, Pukateria, and Bursinopetalum) are not recognized by me as Sederacece. He proposes two new genera (Pseudopanax and Tetrapanax), which I have adopted. In 1863 I pointed out, in the 'Journal of Botany,' that Horsfieldia, Astrotricha, Hydrocotyle (in part), etc., hitherto placed in Umhelliferce ON THE POLYANDROUS GENERA. 3 must be refeiTed to Hederaceee, on account of their ti-uly valvate pe- tals; and in 1864 I added Crithmum to the list. I. On the Polyandeous Genera. In the 'Botauical Magazine' for April, 1856, there is published, on plate 4908, a new genus oi Hederacece, Tupidanthus calyptratus. Hook, fil. et Thorns., of which it is remarked, that " the coalescence of the calyx lobes and corolla into an arched coriaceous calyptra, together with the numerous stamens, the total absence of styles, and very numerous cells of the ovary, are perhaps unique in the Order" to which it belongs. The characters are certainly very singular, but it had evidently been overlooked that two years previously (in 1854) Asa Gray described two allied genera from the South Sea Islands (Plerandra and Tetraplaaandra), both of which share with Tupidanthus a calyptrate co- rolla, polyandrous stamens, and a many-celled ovary. The calyptrate corolla had previously been noticed in some of the older genera of the Order, but the polyandrous stamens were certainly quite a new feature, — no more 13 than had been known to exist amongst this group of plants. During my exploration of the Viti Islands, I was fortunate enough to dis- cover several additions to the polyandrous Hederacea, among them two entirely new genera (Nesopanam and Bakerid), and in the following pages I propose to give a description of them, together with an enu- meration of all the polyandrous Hederacete known to me. In this list a polyandrous genus will be missed, which being referred by Bentham and Hooker fil. (Gen., Plant, p. 17) to Hederaeed, — "est Araliacea anomala ovario subsupero " are their exact words, — would naturally be looked for in this place. I mean Trochodendron, a Japan genus founded by Siebold and Zuccarini, and placed by them amongst Winterea, by Endlicher as an anomalous genus at the end of Magnoliacees, and by Miers in the neighbourhood of TernstrcemiacesB. However it cannot be admitted amongst Hederacea, differing as it does from all the known members of the Order by its 4-celled anthers and many-ovuled ovaries, to say nothing of its entire want of calyx and corolla. Its afSni- ties are, in my opinion, much more with Wintered, the very group in which Siebold and Zuccarini placed it ; and its nearest ally I hold to be Euptelia, Sieb. and Zucc, the nature of which has become better understood by the recent publication of Drs. Hooker arid Thomson's new Indian species ('Linnean Journal,' sect. Botany, vii. 241. t. 2), B 3 4 REVISION OF THE NATURAL ORDER HEDERACE^. in which it is convincingly shown that the best place to be found for that genus is with or near Winterece. This is exactly the position in which Trochodenihon has been put by its founders. The two genera agree in almost all essential points, viz. their arboreous habit, alter- nate, serrate simple and exstipulate leaves, peculiar process of bud- ding, apetalous flowers, numerous stamens, 4-ceUed anthers, many- ovuled carpels (at least in two out of the three known species), ana- tropous ovules, albuminous seed, and minute embryo ; but they differ in the carpels being separate in Euptelia, and tilmost consolidated in TrocJwdendron. How far this woidd affect their being grouped together is a matter of individual opinion. In Sanunculaeea we have genera with free, and more or less consolidated carpels, and in Fa- pavei'acea a species, the common garden Poppy, one variety of which* has distinct and consolidated carpels in the same flower. I incline to think that we have here the first known members of a new Order of plants which will have to be ranged near Rcmunculacece and Magnoliacea, and which might be called Ti'ocIiodendrecB. The two genera I have to propose (Bakeria and Nesopanax) present a new feature, combining as they do free petals with an indefinite number of stamens, the other polyandrous genera previously known having calyptrate petals. " CONSPECTUS GENERUM HEDEBACEAEUM POLYANDE.AEUM. * Fetala calyptraivm cohcBrentia, eaduca. 1. TupidantJius. Pedicelli inarticulati. Plores ecalyculati. Petala 5. Stamina indeflnita, biserialia. Ovarium multiloculare. Stigma depres. sum, 3-4-crure. — Arbor Indies orientalis, exstipulata, foliis digitatim 8-9-foliolatis, umbellis compositis. 3. Tttraplasandra. Pedicelli inarticulati. Flores ecalyculati. Pe- tala 7-8. Stamina 28-32, uniserialia. Ovarium 7-10- loculare. Stigma obsolete 7-10-rafliatum, stylopodio brevi impositum. — Arbor Hawaii- ensis, exstipulata, foliis pinnatim B-7-foliolatis, umbellis paniculatis. 3. Flerandra. PediceUi inarticulati. Flores ecalyculati. Petala 5. Stamina indefinita, pluriserialia. Ovarium 14-15-loculare. Stigma truncatum, obsolete multiradiatum, stylopodio conico impositum. — Arbores Vitienses, exstipulatoe, foliis digitatim 9-foliolatis, umbellis compositis. * See figure iu Card. Chronicle for 1859, and ' Bonplandia,' 1859, p. 336. ON THE FOLYANDROUS GENERA. 5 4. Brassaia. Pedicelli subnulli. Flores calyculati. Petala 7-17, viilgo 12. Stamina 7-17, plerumque 13, uniserialia. Ovarium 7-17, vulgo 13-loculare. Stigma obsolete 7-17-i'adiatum, stylopodio conico impositum. — Arbores Novae HoUancliee, Novae Guineai, et Sumatrae, stipulatae, foliis digitatim 9-16-foliolatis, capitulis raoemosis, pedicellis basi bracteolatis,, „ 5. Reynotdsia. Pedicelli inarticulati. Plores ecalyculati. Petala 8-10. Stamina 8-18, uniserialia. Ovarium 8-18-loculare. Stigma 8-18-radiatum, stylopodio conico impositum. — A.rbores insularum Oceani Pacifici, exstipulatae, foliis pinnatim 3-9-folio!atis, umbellis paniculatis compositis, pedicellis nudis. ** Petala basi coheerentia, persistentia. 6. Gastonia. Pedicelli inarticulati. Flores ecalyculati. Petala 10-11. Stamina petalorum nuraero, uniserialia. Ovarium 8-IO-I0- culare. Styli 8-10, distinct!. — Arbor Mauritiana, foliis exstipulatis imparipinnatis, umbellis paniculatis. 7. Grotefendia. Pedicelli articulati. Flores calyculati. Petala 11-13. Stamina petalorum numero, uniserialia. Ovarium 5-10-locu- lare. Styli 5-10, distincti. — Arbores Mauritianae, foliis exstipulatis imparipinnatis, ijoribus racemosis v. umbellatis. *** Petala libera, persistentia. 8. Bakeria. Pedicelli inarticulati. Flores ecalyculati. Petala 5. Stamina 15, imiserialia. Ovarium 5-loculare. Stigma obscure 5-fidum, stylopodio subconico impositum. — Arbor Vitiensis, foliis digitatim 5-foliolatis, petiolis basi stipulatim dilatatis, umbellis compositis. 9. Nesopanax. Pedicelli inarticulati. Flores ecalyculati. Petala 5. Stamina indefinita, pluriserialia. Ovarium 5-7-loculaTe. Styli 5-7, distincti, breves. — ^Arbor Vitiensis, foliis digitatim 7-9-foliolatis, pe- tiolis basi stipulatim dilatatis, uriibellis compositis. I. TupiDANTHtJS, Hook. fil. et Thoms. Pedicelli inarticulati. Flores ecalyculati, hermaphroditi. Calj'X tubo late clavato v. hemi- spbaerico, limbo truncato. Petala 5, calyptratim cohserentia (" corolla monopetala," C. Koch). Stamina indefinita, biserialia ; antherae oblon- gae, 2-loculares. Ovarium multiloculare, centro depressum et rima stigmatifera 4- v. 3-cruri insculptum, loculis uniovulatis. Bacca co- riacea, multilocularis, loculis 1-spermis. — Arbor Indiae orientalis, foliis exstipulatis digitatim 8-9-foliDlatis, foliolis petiolulatis oblongis acu- 6 EEVISION OF THE NATUKAL OEDEK HEDEEACE^. minatis, integerrimis, glaberrimis, umbeUis compositis, lateralibus, eorol- lis viridibus, staminibus pallide stramineis. — Ta- pidanthus, Hook. fil. et Thorns, in Bet. Mag. t. 4908 (1856) ; C. Koch, Wochenschrift, 1859, p. 348. 1. T. calffptratuStHodk. fil. et Thoms. inBot. Mag. t. 4908. T. Piickleri, C. Koch, in Wochenschrift, 1859, p. 348, cum icon. Sciadophyllum pulcJiellum, Hort. — At the base of the Khasia Mountains, East- ern Bengal (Hooker fil. and Thomson ! -Griffith ! Tupidanthns calyptratns {partly after SooJcer). Jl- *701). This tree seems to grow as underwood and support its weak stem by leaning against other trees, as many other EederacetB do, a habit which the founders of the genus expressed by "Arbor alte scandena.*' Dr. Carl Koch, mistaking "scandens" for " cirrhosus" {"rankend" is his exact translation), has endeavoured to make a second species {T. Puckleri), which he says differs from T. calyptratua by not being cirrhose. Of course T. PUckleri must fall to the ground. The Kew plant, from which the figure in Bot. Mag. t. 4908 was made, is now 24 feet high, and erect, in growth very much the same as all the other garden plants of this species are. II. Tetkaplasandka, a. Gray. Pedicelli inarticulati. Elores eca- lyculati, polygami (?). Calyx tubo hemisphserico, limbo brevissimo truncato vix denticulato. Petala Unearia, 7-8, calyptratim cohaerentia. Stamina petalorum numero 4-plo (38 v. 33), uniserialia; antheree ob- longse, subsagittatse. Ovarium 7-10-loculare. Stigma obsolete 7-10- radiatum, stylopodio brevi conico impositum, loculis 1-ovulatis. Dmpa baccata, 8-10-pyrena, pyrenis coriaceis. — Arbor Hawaiiensis, procera, inermis, foliis exstipulatis pinnatim 5-7-foliolatis, foliolis oblongis v. ellipticis utrinque obtusis v. apice acutis integerrimis, subtus incanis. ON THE POLYANDROUS GENEKA. umbellis paniculatis. — Tetrapldsandra, A. Gray Bot.Wilkes.p. 727,t. 94 (1854). Walp. Ann. v. p. 82. 1. T. Hawaiensis, A. Gray, Bot. Wilkes, p. 728, t. 94 ; Walp. Ann. V. p. 83. — Hawaii, Sand- wich Islands, in the dis- trict of Puna (Unitt-d States Expl. Exped.). III. Pleeandea, a Gray. Pedicelli inarticii- lati. riores ecalyculati, polygami. Calyx tubo turbinate, limbo brevis- simo post anthesin re- pando-undulato. Petala ovato-triangularia, 5, ca- lyptratim cohserentia. Stamina indefinita, pluriserialia ; anthevse oblongse. Ovarium 12-15-\ocu- lare, loculis 1-ovulatis. Stigma truncatum v. depressum, obso- .lete multiradiatum, stylopodio conico impositum. Drupa obo- vata, 12-15-loculare, loculis 1- spermis. — Arbores Vitienses, inermes, foliis exstipulatis digi- tatim 9-folioIatis, foliolis obo- vato-oblongis integerrimis, um- bellis compositis. — Plerandra, A. Gray, Bot. Wilkes, p. 729, t. 95 (1854) ; Walp. Ann. v. 81. Asa Gray not having had suf- iiclent materials, I have emen- dated the generic character of Plerandra. The petals are 5, Tetraplasandra Hawaiensis {after Asa Gray). Pleraadra Grayi. REVISION OF THE NATURAL ORDEE HEDERACEJ!. and calyptrate. Hitherto there was only one species known ; the second, which. I was enabled to add, I have named in honour of the illustrious founder of the genus, P. Orayi. 1. P. Pickeringii, A. Gray, Bot. Wilkes, p. 729, t. 95 (1854) Walp. Ann. v. 81. — Viti Levu (Seemann ! n. 306); Vanua Levu, above Nandy (Milne !) ; Ovalau, Viti Islands (United States Expl. Exped.). "A tree, called 'Vola' by the natives of Vanua Levu" (Milne). 3. P. Orayi, Seem. Viti, Appendix, n. sp. ; inermis, glabra, foliis digitatis, foUolis 9 obovato-oblongis obtusis basi in petiolum attenuatis integerrimis, umbeUis multiradiatis, umbeUulis 36-floris, calyce post anthesin repando-undulato, petalis 5 valvatis oblongis acutis calyptra- tim cohserentibus, mox deciduis, staminibus indefinitis, stigmate ob- scure multiradiato, ovario 13-15-loculari, drupa obovata obscure 13- 15-costata. — Viti Levu, Viti Islands (Seemann!, n. 308). A small tree. Petioles 13 inches, petiolules 1|- inch long. The upper leaflets the largest, and their blade 6-7 inches long and 3^ inches broad. Plowers greenish. Drupe about f of an inch long, -j inch in diameter. Stigma depressed in the male flowers, on conical stylopodla in the hermaphrodite. IV. Brassaia, Endl. Pedicelli subnulli. Flores calyculati, calyculo 4-phyllo, polygami. Calyx tubo obconico, limbo truncato. Petala linearia, 7-17, plerumque 13, calyptra- tim cohserentia. Ovarium 7-17-locu- lare, vertice exserto multicostatum, loculis 1-ovulatis. Stigma sessile, 7- 17-radiatum, stylopodio conico imposi- tum. Drupa angulato-sulcata, 7-17- vulgo 13-locularis. — Arbores inerraes, Novam HoUandiam, Novam Guineam, Amboynam et Sumatram incolentes, stipulate, foliis digitatim 7-16-folio- latis, foliolis subiutegerrimis, capitu- lisracemosis, pedicellis subnuUis. — Brassaia, Endl. Nov. Stirp. Mus. Vin- dob. Dec. n. 100; Gen. Plant. Suppl. i. p. 1415. Parapanax, Miq. PI. Ne- derl. Ind. Suppl. i. p. 339, ex parte. Brassaia actinophylla. SciadophylU, Paratropia sp. auct. ON THE POLYANDROtJS QENEKA. 9 The genus Brassaia, though well defined by Endlicher, has not been recognized by other botanists. Several years after its publication (1843) Bentham described a new Hederacea from New Guinea, which he referred to Sciadophyllum, and which, in 1856, was transferred to Paratropia by Miquel, though he liad not seen the plant ; and, iii order to admit it into that genus, had to amplify the generic character. More recently the same author, stiU unaware of the existence of Bras- saia, established the genus Parapanax upon two Sumatran species, with- out, however, suspecting that his Paratropia macrostacltya was con- generic with one of them. It is to be hoped that, in future, this truly natural genus wiU be more generally recognized. We now know repre- sentatives of it from tropical New Holland, New Guinea, Amboyna, and Sumatra, and the intermediate regions will doubtless supply addi- tional species. 1. B. actinophylla, Endl. Nov. Stirp. Mus. Vindob. Decad. i. p. 89 ; Walp. Kep. i. p. 430, v. p. 935 ; E. Mueller, Eragmenta, ii. p. 108.— Tropical parts of Eastern Australia (Sir Joseph Banks ! A. Cunning- ham 1 n. 484, of Second ' Mermaid's ' Voyage in Mus. Brit., Eobert Brown ! in Herb. Hook.) This fine tree was first discovered by Sir Joseph Banks, next found by Eobert Brown, and afterwards met with " at Pine Head and similar situations " of the east coast of New Holland, north of Endeavour Eiver by A. Cunningham, who, in his journals sent to Sir J. Banks, enume- rates it under n. 484, as an Uriicacea, appending to it the following note : — " Amentum [capitulum] S-lO-flomm, subglobosum, pedicel- latum, pedicello crasso. Elores 13-andri, hermaphroditi. Calyx semi- superus, subcyathiformis, dentibus \i. e. petalis] apice conniventibus deci- duis, 3-bracteatus. Stam. 13, antherifera, calyci inserta. Anth. 3-Io- cul. Ovarium lO-loculare. Stigm. sessile, radiatum. — Arbuscula 16-30 ped. ramis crassis brevibus, foliis 7-12-nis, foliolis ellipticis petiol. glabris raucronatis obtusis." Erom Dr. F. Mueller (Eragm. 1. c.) we learn that Mr. Charles Moore found this tree near Boyn Eiver, Mr. Eug. Eitzalan near Port MoUe, and Leichhardt in those parts of E. Australia which he crossed. By the last-named author the tree is stated to be forty feet high, and the leaflets varying in number to sixteen. According to Ch. Moore the flowers are said to be scarlet, but this statement is hot confirmed from what is known from other sources, nor borne out by an inspection of the specimens before me. In the copy of 10 REVISION OP THE NATURAL ORDER HEDEEACEiE. ' Endlicher's Tconographia" which I have consulted, there is no plate of this Brassaia, though Endlicher quotes it himself; and G. Koch says (' Wochenschrift,' IS 59) that it has never been published. 3. S. macrostachya. Seem. mss. — Sciadophyllum macrostachywm, Benth. in Lond. Journ. of Bot. ii. (1843), p. 222 ; Walp. Kep. ii. p. 939. Paratropia macrostachya, Miq. in Bonplandia, 1856, p. 139 . — New Guinea (Hinds! in Herb., Benth.). 3. B. littorea. Seem. — Papaya littorea seu Papaja Pante, Eumph. Amb. i. p. 150. t. 52. — Halong, Amboina, rare (Eumphius). Leaves over 3 feet long, leaflets 13-16 elliptico-lanceolate, 12-15 inches long, 3 inches broad, acute on both ends. Eacemes paniculate, over a foot long. Called by some Amboinese " Lau Takka," from the resemblance of its leaves to those of Tacca. Does not seem to have been collected since the time of Eumphius. 4. B. sessilis, Seem. mss. — Parapanax sessile, Miquel, Fl. Nederl. Ind. Suppl. i. p. 339. — Western Sumatra, near Lolo (Teijsmann). Species exclusa : — B. palmata, Dene, et Planch. =2V'e»e«2fl Sundaica, Miq. fide C. Koch. V. Eeynoldsia, a. Gray. Pedicelli inarticulati. Flores ecalycu- lati, polygami. Calyx tubo oboonico, limbo integerrimo v. subre- pando. Petala linearis, 8-10, calyptratim co- hserentia. Sta- mina 8-10, uni- serialia ; antherae lineares. Ova- rium 8-I8-I0- culare. Stigma 8-18-radiatum, stylopodio co- nico impositum. Drupa baccata, globosa, 8-18- pyrena, pyrenis cartilagineis. Embryo minu- Reyuoldsia Sandwicensis {after A. Grafjr. ON THE POLYANDROTJS GENERA. 11 tus, radicula supera cylindrica. — Arbores insularura Oceani Pacified, in- ei'mes, glabrse, exstipulatse, foliis pinnatim 3-9-foliolatis, foliolis ovatis V. suboordatis detvtatis, umbellis paniculatim compositis. — Reynotdsia, A. Gray, Bot. Wilkes, p. 723, t. 92 et 93 ; Walp. Ann. v. p. 82. 1. R. Sandwicenm, A. Gray, Bot. Wilkes, p. 724, t. 92; Walp. Ann. V. p. 82. -^In a ravine near Waianse, Oahu, Hawaiian (Sand- wich) Islands. 3. R. pleiosperma, A. Gray, Bot. Wilkes, p. 725 ; Walp. Ann. v. p. 82. — forests of Savaii, Samoan (Navigator) Islands. 3. Reynoldsia verrucosa. Seem. n. sp. ; foliis pinnatis, foliolis 3-4- jugis cum imparl ovatis acuminatis dentatis, dentibus glanduloso-ver- rucosis, umbellis 5-7-floris, pedieellis ancipitibas, calyce undulato-mul- tidenticulato, corolla clausa subobovata, petalis staminibusque 7, ovario multiloculari, stigmate multiradiato, fructu . . . — Tahiti (Nelson ! col- lected in Captain Cook's third voyage). Differs from the allied R. pleiosperma of the Samoan Islands in the quite warty teeth of the leaves, and compressed pedicels. The whole plant glabrous. Leaflets petiolulate, 2^3 inches long, 1^ inch broad. The umbels axillary, shorter than the leaves. All the flowers in Nelson's specimens herma- phrodite. VI. Gastonia, Comm. Pedicelli inarticulati. Plores ecalyculati, hermaphroditi. Calycis tube obconico, angu- lato. Umbo subintegro. PetaJa 11, linearia, 1-nervia, basi plus minus cohserentia, ses- tivatione valvata. Stamina petalorum numero sequalia; antherse oblongse. Styli 8- 10, basi ima comiati, erecti, demum recurvi. Drupa ob- longa, 8-10-costata, 8-10- locularis. — Arbor Mauritiana, foliis imparipinnatis, foliolis 5 integerrimis, umbellis pa- niculatis, paniculis termina- Gastoaia cutisponga. 12 KEVISION or THE NATURAL OKDEB HEDERACEjE. libus. — Gastonia,CommAn Juss. Gen. 317 ; Lam. Diet. ii. p. 610 ; De Cand. Prodr. iv. p. 356. The genus Qastonia has been incorrectly described. De Candolle C Prodi-omus '), to go no further back, assigns to it from 5-16* petals, and double the number of stamens, two stamens being said to be placed before each petal. Decaisne and Planchon describe the petals as 3- nerved. All three are wrong. The stamens in this, as in all other genera of the Order with definite stamens, alternate with the petals, and are equal in number with them; and they are always 1 -nerved. What has given rise to the belief of their being 3-nerved is that they are very narrow, and are more or less coherent, in some instances not separating at all from each other, so that the corolla appears to have fewer petals than it really has. The normal number of petals and carpels seems to be 10. But there are' sometimes fewer by abortion, or more by excess. Oastonia is closely alKed to Orotefendia, a genus also inhabiting the islands of Eastern Africa. Both have imparipinnate leaves and very narrow petals, but they are easily distinguished from each other, viz. : — Gastonia. PedicelH inarticulati. Plores ecalyculati. Orotefendia. Pedicelli articulati. Flores calyculati. 1. G. cutuponga. Lam. Diet. ii. p. 610 ; De Cand. Prodr. iv. p. 356 ; Q. spongiosa, Pers. Ench. ii. p. 30 ; Bois A'iponge, incolarum. — Mauritius (Sieber! PI. Maurit. Exs. ii. p. 197, Carmichael! Bouton! in Herb. Kew.), growing in woods, abundantly in Mount "Pouce." 3. Speoies dubia, indescripta : — Q. saururoides, B.oxh. Cat. Hort. Calc. p. 70; Gilibertia saururoides, De Cand. Prodr. iv. p. 356, in the Moluccas. Species excluaee : — G. aculeata, Hortoi-.^ Cupkocarpus acukatus, Dene, et Planch. G. Candollei, Rortor. =■ Brasaaiopsis speeiosa, Dene, et Planch. G. dentata, Hortor. = Brassaiopsis speeiosa, Dene, et Planch. G. longifolia, Horior. =Brassaiopsis speeiosa, Dene, et Planch. G. Nalugu, Lam. (Gilibertia Nalugu, De Cand.)=Zeefl staphyka, Eoxb. teste Wight et Am. Prodr. ;=£. sambucinn, Willd. teste Miq. G. (?) Oahmnsis, A. Gray, Bot. Wilkes, p. 736, is the type of a new genus. G.palmata, TS,oyih.= Trevesia palmata, Visian. * " 16 " may possibly be a misprint for " 6," else the number would be 32. ON THE POLVANDUOUS GENERA. 13 Grotefeodia cuneata. VII. Grotefendia, Seem. n. gen.. Pedicelli articulati. Elores calyculati, hermaphroditi. Calyx tubo oboonico angulato, limbo ob- scure 5-6-lobato V. sub- intcgerrimo. Petala cu- neato-liuearia, 1-nervia, 11-13, basi cohffirentia, sestivatione valvata. Sta- mina 11-13; anthersB 11- neares. Pollen ellipti- cura. Styli 5-10, com- pressi, erecti v. demum recurvi. Ovarium 5-10- loculare, loculis 1-ovula- tis. Dnipa oblonga, coa- tata V, sublsevis. — Arbores Mauritianse, foliis exstipu- latis, imparipinnatis, fo- liolis integerrimis v. ob- solete dentatis, floribus umbeUatis v. racemosis, umbellis v. racemis paniculatis. This genus I have named in honour of Dr. Grotefend (one of the earliest decipherers of cuneate writings, and Director of the college in which I was educated), and of his son, to whom I am indebted for my first lessons in botany. 1. G. cuneata. Seem. mss. Glabra ; foliis imparipinnatis, foliolis 11 longe petiolulatis ovatis v. subrotundato-ovatis obtusis basi acutis V. obtusis, penninemis, subrepando-dentatis v, integerrimis, floribus racemosis, pedicellis subnuUis, calyculo 5-dentato, calycistubo obconico, limbo subintegro, petalis cuneato-linearibus 18 1-nerviis plus minusve coalitis, staminibus 13, antheiis linearibus, pollinis granulis ellipticis, stylis 10 basi laeviter cohserentibus, ovario 10-loculari, loculis 1-ovulatis. — Gastonia spongiosu. Herb. Hook, non Pers. — Mauritius (Carmichael ! in Herb. Hook. ; Hardwicke 1 in Mus. Brit.). Leaflets 3-3|- inches long, Zy- 3 inches broad ; petiole |-f inch long. 3. G. paniculata. Seem. mss. Glabra ; foliis imparipinnatis, foliolis 7 oblongis elliptico-oblongis acutis v. obtusis basi acutis integerrimis penninerviis, floribus racemosis, pedicellis articulatis brevissimis, calyculo 5-dentato, calyois tubo obconico, 5-6-angu- 14 REVISION OP THE NATURAL OKDEK HEDERACE^E. lato, limbo B-6-dentato, petalis cuneato-Unearibus 11 l-nerviis, plus minusve cohaerentibus, stamiiiibus 11, stylis 5-7 compressis erectis demum recurvis, fructu oblongo 5-7-angulato S-V-loculari. — Gastonia heplaffyna, Herb. Hook. ; Oilibertia paniculata, De Cand. Prodr. iv. p. 256. — Mauritius, in mountainous forests (Carmichael ! Bouton ! in Herb. Kew.). Leaflets 6-7 inches long (or smaller), 3-3 J inches broad. That Giliiertia paniculata, De Cand., described from a scrap, is a synonym of this species, I was able to determine from a few flowers kindly sent to me by M. Alph. de Candolle for that purpose. 3. O. repanda, Seem. — Oilibertia repanda, De Cand. Prodr. Iv. p. 256.— Mauritius (Bory ! in Herb. De Candollei). Easily distinguished from the foregoing species by its umbellate flowers and long pedicels. I am indebted to M. Alph. de Candolle for a sight of a few flowers of this species, sufiicient for determining the genus. VIII. Bakeria, Seem. mss. n. gen. Pedicelli inarticulati. Flores ecalyculati. Calyx tubo tm'biuato cum ovario connato, limbo brevis- sirao repando-undulato. Petala 5, ovato-triangularia, sestivatione val- vatct, apice incurvula, libera, intus 1- I nervia. Stamina 15, uniserialia ; fila- menta compressa; antherse oblorigee. Ovarium 5-loculai-e, loculis 1-ovula- tis. Stigma obscure 5-fidura, stylo- podio subconico brevi 5-angulato 6ul- cato impositum ; fructu . . . — ^Arbor Vitiensis mediocris, glaberrima, in- ermis, foliis digitato-5-foIiolatis, fo- liolis longe petiolulatis obovato-ob- tusis V. acutis in petiolum angustatis integerrimis, petiolo basi stipulato- dilatato (ut in Agalma), umbellulis umbeUatis exinvolucratis, floribus viri- Bakeria Vitiensis. diusculis. This new genus, named in honour of Mr. J. G. Baker, of Thirsk. Yorkshire, a distinguished British botanist, difl'ers from Flerandra in having free petals, a definite number of stamens (15) arranged in a ON THE POLYANDROUS GENEEA. 15 single series, and ^ 5-celle(l ovary ; from Teiraplasandra, in having only 5 petals and a 5-celled ovary, and a different habit; from Rey- noldda, in having three times as many stamens as petals, and a 5-ceIIed ovary ; and from all the other genera of the Order in having 5 free petals, 1 5 stamens, and a 5-celled ovary. There is only one species, viz. : — 1. B. Vitiensis, Seem. n. sp. ; a slender tree ; petioles 4-5 inches long, petiolules 1 inch long; blade of leaflets 3-4 inches long, 1^-3 inches broad ; pedicels 4-angular, not articulated. — Naraosi, interior of Viti Leva (Seemann!, 209); also cCllected in the same island (foliage only), twenty miles inland, and there plentiful, by Milne. IX. Nesopanax, Seem. n. gen. Pedicelli inarticulati. Flores ecaly- culati. Calyx, tubo obconico, cum ovario connato; limbo supero, obsolete 5-dentato. CoroUae petala 5, ovato- triangularia, disci epigyni margini inserta, libera, aestivatione valvata. Stamina indefinita, cum petalis in- serta, pluriserialia; iilamenta brevia ;, antherae oblongse. Ovarium inferum, 5-7-loculare. Ovula in loculis so- litaiia, pendula. Styli 5-7, breves, distincti ; stigmata simplicia. Drupa oblonga, 5-7-locularis, calycis limbo stylisque coronata. — Arbor Yitiensis mediocris, glabra, inermis, foliis digi- tatis, foliolis 7-9 obovato-oblongis utrinque attenuatis integerrimis, pe- tiolis basi stipulate dilatatis, um- bellis multiradiatis, umbellulis 26- 30-floris, floribus viridibus. This genus differs from Plerandra, A. Gray, in having free petals, 5-7 distinct styles, and a 5-7 -celled drupe. It agrees with it in habit, and tlie indefinite number of sta- mens. I have only one specifes : — 1. N. Vitienm, Seem. n. sp. — Viti Islands (Seemann ! n. 207, Milne!). Nesopanax Vitiensis. 16 HEVISION OP THE NA.TUEAL ORDEE HEDEEACE^. Petiole li feet long. Leaflets pinnately veined, coriaceous ; blade 6-7 inches long, 2^3 inches broad; petiolules 1^-2 inches long. Peduncles 6-8 inches long. Pedicels of fruiting specimens l|-2 inches long. There is no ripe fruit. II, On the GrENEEA WITH A SINGLE StYLE. There are only a few genera having a single style, genera in which the ovary is elongated into a short conical stylopod being more com- mon; these are Brassaiopsis, Macropanatx, Pentapanatc, Agalma, Bendropanax, and Hedera. I follow Lowe in his ' Flora of Madeira ' in calling the style single rather than simple, because in Hederaeece that organ is always composed of two or more styles, in the genera here pro- visionally grouped together so closely united as to form a single column, and separating either not at all or just a little at the top ; but they never become disunited to the very basBj and bend back as in Gastonia, Foh/scias, Gilibertia, etc. CONSPECTUS GENEBTJM HEDEEACEAEUM STYLO UNIOO. * Ovarium 2-locwlare. 10. Brassaiopsis. Pedicelli inarticulati. Mores ecalyculati. Al- bumen sequabile. — Arbores aculeatse Asise tropicse et subtropicse, foliis simplicibus, palmatis v. digitatim sectis. W. Macropanax. Pedicelli articulati. Hores calyculati. Albumen ru- minatum. — Frutices inermes ludise orientalis, foliis digitatim compo- sitis. ** Ovcmum h-{^er excessimi G-)locnlare. 12. Fentapanax. Pedicelli articulati. Flores calyculati. Drupa cxsucca. Albumen eequabile. — Arbusculae inermes Indiae orientalis, foliis pinnatis. 13. Agalma. Pedicelli inarticulati. Flores ecalyculati. Drupa exsucca. Albumen sequabile. — Arbores inermes Asise tropicse, foliis digitatim compositis. 14. Bendropanax. Pedicelli inarticulati. Flores ecalyculati. Drupa baccata. Albumen sequabile. — Arbores inermes Americas et Asise tropicse et subtropicse, foliis simplicibus. 15. Hedera. Pedicelli inarticulati. Flores ecalyculati. Drapa ON THE GENERA "WITH A SINGLE STYLE. 17 baccata. Albumen ruminatum. — Prutices sarmentosi, inermes, in Europa Asia et Africa indigenee, foliis simplicibiis. X. Brassaiop- sis, Dene, et Planch. Pedicelli inartica- lati. Flores ecaly- culati, hermaphro- diti. Calyx tubo obconico, limbo 5- dentato. Petala 5, VV"'1«'> ovato - triangularia, 1-nervia, apice in- curvula, libera, ses- tivatione valvata. Brassaiopsis speciosa {partly after Soo&er). Stamina 5, petalis alterna ; filamenta filiformia ; antherse subrotundatse. Ovarium 2-loculare, loculis 1-ovulatis. Ovula pendula. Stylus uni- cus, compressus; stigma obscure 2-lobatum. Drupa subgiobosa, 2- looularis. Semina solitaria. Albumen sequabile. — Arbores aculeatse in Asia tropica et subtropica vegetantes, stipulis nullis, foliis alternis longe petiolatis palmato-5-lobis, lobis argute dentatis v. digitatim 5-8-sectisf segmentis dentatis, umbellis globosis longe pedunculatis in racemes v. paniculas dispositis involucratis, pedunculis pedicelHsque compressis, floribus albidis. — Brassaiopsis, Dene, et Planch. Kev. Hort. 1854, p. 106 ; Planch. Hort. Donat. p. 8 ; Hederee, Panacis, Ka- lopanacis et Aralice sp. auct. C. Koch (' Wochenschrift,' 1859, p. 364) has endeavoured to show that Brassaiopsis and Macropanax are identical, but he had evi- dently overlooked that Brassaiopsis (of which Hedera glomerulata, Bot. Mag. t. 4804, is the type) has inarticulate pedicels, no "caly- culus, and a very distinct habit. It is true that Miquel referred Hedera glomerulata to his genus Macropanax ; nevertheless it does not agree with the generic character he furnishes, and is very different from M. oreopMlum. Decaisne and Planchon, ia their valuable paper on HederaoecB, ascribe to their new genus Brassaiopsis two short styles, but there is only one, as in the absence of specimens may be seen rom Hooker's very correct figure; and by a slip of the pen the sland of Bourbon, instead of Java, is named as the native country of B. speciosa. c 2 18 KEVISTON OF THE NATURAL ORDER 1IEDERACEJ3. 1. B. Hainla, Seem. rsni.—Hedera Hainla, Hamilton, mss. in Don, Prod. Fl. Nap. p. 187 (1835) ; De Cand. Prodr. iv. p. 253. H.poh/a- cmtla, Wall. Cat. n. 4907 ; Wall. Plant. Asiat. t. 190 ; Don, Gen. Syst. iii. 392 ; Walp. Eep. ii. p. 431 and 432. E. scandens, De Cand. Prodr. iv. p. 264 ? Fanax palmatum, Eoxb. H. Ind. ii. p. 74 (1832). P. ciircifoUa, Griff. Itineraiy' Notes, p. 145. Aralia palmata. Lour. Fl. Cochinch. p. 233 ? non Lam. J. scandens, Poir. Suppl. i. p. 419? Nepal (Hamilton ! in Mus. Brit. ; Wallich ! n. 4907) ; moist valleys of CWttagong (Roxburgh) ; Bootan (Griffith ! n. 2667), towards Oon- gar Bridge, in forests, 5-6000 feet (Griffith! n. 697), Sikkim (Hook, fil. et Thoms. !). 3. B. ricinifolia, Seem. mss. — Panax ricinifolium, Sieb. et Zucc. in Abhandl. Bair. Akad. iv. 2, p. 198. — Kalopanax ricinifolium, Miq. Ann. Mus. Bot. Lugd.'Bat. vol. i. p. 16. Japan (Siebold ! in Herb. Benth.), N. China (Fortune ! in Herb. Mus. Brit.). 3, B. Uspida, Seem. sp. n. ; ramis petiolis pedunculisque hispido- aculeatis, aeuleis rectis, foliis palmato-10-lobis, lobis ovatis v. oblongis acuminatis setoso-serratis, supra glabris, subtus ad nerves ferrugineo- hirtellis hinc inde aculeatis, umbellis multifloris (50-60) in paniculas terrainales dispositis, pedicellis rufo-hirtellis, calycibus tomentosis, pe- talis glabris.— Bootan (Griffith! n. 2066 in Mus. Brit.). Judging from the dried specimen, the peduncles are stiff and patent, not drooping as in B. speciosa. Leaves nearly a foot across ; peduncles 5 inches long ; bracts surrounding the base of the pedicels lanceolate- linear acute ; the whole inflorescence forming terminal panicles l-l-j feet high. 4. B. confluens, Seem. ; ramis petiolisque aculeatis, foliis glabris pal- matim 8-9-lobis, lobis pinnatiiidis v. bipinnatifidis, umbellis globosis longe pedunculatis solitariis v. racemosim dispositis, junioribus stellato- pubescentibus, calyce 5-d6ntato, stylo 1, fructu subgloboso* latiore leviter compresso, 2-spermo. — Hedera confluens, WaU. Cat. n. 4910, ex parte; Nepal (Wallich ! n. 4910, ex parte). The sheet of Wallich's Herbarium at the Linnean Society containing this plant, has two other Hederacene pasted on it, all three of which are labelled n. 4910. One of them is probably another species of Brassaiopsis, with unarmed petioles and palmate 3-5-lobed leaves ; the other is a Trevesia. ON THE GENERA WITH A SINGLE STYLE. 19 The leaves of B. confiuem are \\ foot across. The petiole is at the top expanded into the blade, from which spring 8-9 mostly bipinnatifid lobes. It is a magnificent species. 5. B. sjieciosa, Dene, et Planch. 1. c. — Macropanaas glomerulatum; Miquel in Bonplandia, 1856, p. 139. Aralia glomernlata, Blum. Bijdr. p. 873. Hedera glomerdata, De Caad. Prodr. iv. p. 265 ; Hook. Bot. Mag. (1854) t. 4804; Kegel, Gartenflora (1863), tab. 411. Bras- saiopsisfloribunda, Dene, et Planch. Hort. Donat. p. 8, excel, syn. Wall. Gasionia longifolia, hortorum. — Java (not Bourbon, as stated liy Dene, and Planch.). Decaisne and Planchon have introduced considerable confusion into the synonymy of this species by abandoning (Hort. Donat.) the name speoiosa, which a few months previously they gave to the plant, and quoting Wallich's H. floribunda as a synonym. I have retained the oldest name. 6. B.Jloriiunda, Seem. mss. in Herb. Mus._Brit. — Hederajloribunda, Wall. Cat. n. 4912 A; G. Don, Gen. Syst. iii. p. 394. Hedei-a WallicJiiana, Steudl. Norn. Bot. — Nepal (Wallich ! n. 4912a); Khasia Mountains, 2000-4000 feet high (Hooker fil. et Thomson !) Segments of leaves ovate at base, minutely dentate, or almost quite entire at the edge. Peduncles unarmed. 7. B. aculeata. Seem. mss. in Herb. Mus. Brit. — Hedera aculeata, Ham. in D. Don, Fl. Nepal, p. 185 ; De Cand. Prodr. iv. p. 264. — Nepal (Hamilton! -in Mus. Brit.). Segments of the leaves acute at base, and coarse]y sej'rated at the edge ; umbels on stiff peduncles 2-3 inches long. Obs. — The leaves of the last three species seem, at first sight, com- pound, but on closer inspection it will be found that there is no articu- lation at the base of the apparent leaflets, but that the top of the pe- tiole is expanded, then suddenly contracted, and again expanded into leaflet-like segments. In B. confluens the same structure is repeated on a much larger scale. I took some pains to investigate this struc- ture, as Brassaiopsis seemed to be the only genus of the Order in which simple and compound leaves occurred together. 8. B. cyrtostyla, 'Sieem.—Macropanax cyrtostylum, Miq. Ann. Mus. Lugd. Bat. i. p. 13.— Sumatra (Korthals). ^ XI. Macbopanax, Miquel. Pedicelli articulati. Flores calyculati, polygami. Calyx tubo obconico, limbo 5-(vel per excessum 6-) dentato. Petala 5 v. 6, ovata, libera, sestivatione valvata. Stamina 5 v. 6, petalis 20 REVISION or THE NATURAL OIIPEE, HEDEEACB^. Macropanai nndulatum. alterna; filamenta filiformia; antherae ovatiE. Ovarium 3-loculare, loculis 1-ovulatis. Ovula pendula. Stylus unicus, cylinclricus, stigniatibus 3 or- biculm'i-oonvexis subunitis._ Drupa ellipsoidea, subsicca, laevis, 3-locula- ris. Semina solitaria. Albumen ru- minatum. — Frutices inermes Indise orientalis, foliis exstipulatis digita- tim B-7-foliolatis, foliolis serratis, umbe.llis in racemos v. paniculas dis- positis. — Macropanax, Miq. in Bon- plandia, 1856, p. 139. 1. M. oreophilum, Miquel in Bonpl. , 0. p. 139 ; Mor. Nederl. Ind. vol. i. pars 1, p. 763. — Panax serratum, Wall. ; in De Cand. Prodr. iv. p. 353. Aralia dUperma, Blum. Bijdr. p. 873. Hedera disperma, De Cand. Prodr. iv. 265. Aralia (?) ca- lycnlata, ZoU. et Moritz. Syst. Verz. p. 365. Brassaiopsis disperma, 0. Koch, Wocbenschrift, 1859, p. 364.— Java (Horsfield! in Mus. Brit.; De Vriese! Jungbuhn ! in Herb. Hook.) 3000-4000 feet above the sea; Sumatra and Laronan (fide Miquel), Bootan (Griffith! n. 206); Nepal (Wallich! Cat. n. 4915); Khasia Hills (Hook. fil. et Thomson!). Var, ;8, foliolis multo temtioribus. M. Jloribundum, Miq. 1. c. — Brassaiopsis floribunda, C. Koch, Wochenschrift, 1859, p. 364. — Java (Junghuhn I in Herb. Hook.). 3. M. undulatum, Seem. mss. in Herb. Mus. Brit. — Hedera undulata. Wall. Cat. n. 9416 a; Don, General System, iii. p. 394; Walp. Eep. ii. p. 433 (v. s. sp.)— Silhet (Wallich! n. 9416a); Khasia (Hooker fil. et Thomson) ; Sikhim (Hooker fil. et Thomson!). 3. M. concinnum, Miq. Anu. Mus. Lugd. Bat. i. p. 320. — Java (De Vriese). Species exclusse : itf. cyrtostylum, Miq. = Brassaiopsis cyrtostyla, Seem. M. glomerulatum, Wi.({.-=Brassaiopsis speciosa, Dene, et Planch. XII. Pentapanax, Seem. gen. nov. Pedicelli articulati. Elores calyculati, hermaphroditi. Calyx tubo pbconico angiilato, limbo 5- ON THE GENERA WITH A SINGLE STYLE. 21 dentato. Petala 5, ovata, l-nei-- via, libera, sestivatione valvata. Stamina 5. Stylus 1, 5-angu- latus; stigma 5-lobatum. Pruc- tua oblongus, acute 5-angulatus, B-locularis. — Prutices v. arbus- culae IndiBe orientalis, inermes, foliis alternis, pinnatim 3-5-fo- liolatis, foliolis coriaceis v. sub- membranaeeis, ovatis v. oblou- gis, basi obtusis v. subcordatis, integemmis, serratis v. setoso- serratis, racemis v. umbcUis com- positis, pedicellis ssepe eloiigatis subfiliformibus, apice incrassatis. — Pentapanax, Seem, mss. in Herb, iftus. Brit. Pamcis et Hederce sp. auctor. Pentapanax Lesohenaultii. Flores racemosi. Pedicelli hirtelli P. racemosum. Pedicelli glabri P. Flores umbellati. Pedicelli hirtelli , . P. umb Pedicelli glabri. Foliola coriaeea, iutegerrima .... P. PoUola membranacea, setoso-serrata . . P. . 1. P. racemosum, Seem. mss. in Herb. Mus. Brit. ; glabrum, foliis loiige petiolatis pinnatim 3-B-foliolatis, foliolis petiolulatis obova- tis V. ovato-oblongis, basi obtusis v. subcordatis, acuminatis, sub- integerrimis, floribus distincte raoemosis, racemis longis gracilibus panioulatis, pedunculis pedicellisque liirtellis, pedicellis basi bractea suffultis, calyculo fimbriato (v. s. sp.). Sikhim (Hooker fil. et Thomson! in Herb. Mus. Brit.), 4000-6000 feet above the sea. This plant was distributed by Drs. Hooker and Thomson under Hedera subcordata, Wall., but it is not identical with the species described by Don under that name from Wallich's authentic specimen. The panicles are 1-j- foot long, and at the base (as are the young leal branches) furnished with persistent scales (perulse); 22 REVISION OF THE NATURAL ORDER HEDEKACE*. 2. P. subcordatum., Seem. mss. in Herb. Mus. Brit, ; glabram, foliis pinnatim 5-foliolatis, foliolis longiascule (1-li poll, long.) petiolulatis ovato-o!)longis acuminatis, basi rotundatis vel subcordatis, serratis, ser- raturis obtusiusculis, floribus distincte racemosis, pedunculis pedicel- Usque g;labris, cnlyculo integro. — Hedera suboordata. Wall. Cat. n. 4917; G. Don, Gen. Syst. iii. p. 394 (v. s. sp.).— Silhet (Wallich ! n. 4917, in Herb. Soc. Linn. Lond.) ; Khasia Hills (Hooker fil. et Thomson !). Blade of leaflets from 4-5 j inches long, 3-4 inches broad. Both this and the preceding species diifer in their truly racemose flowers from the other species of Pentapanax, but they agree so well in every other respect, that I cannot separate them generically. 3. P. umbellatum, Seem. mss. in Herb. Mus. Brit., foliis pinnatim 5-foliolatis, foliolis petiolulatis ovatis acuminatis, basi obtusis, serratis coriaceis, floribus umbellatis, urabellis corapositis, pedunculis pedicel- lisque hirtellis, calyculo subintegro (v. s. sp.). — Khasia (Hook. fil. et Thomson !). Dififers from P. LeschemvUii (for which it was mistaken by Hooker and Thomson) in its coriaceous, not setoso-serrate leaves, and hirteUous peduncles and pedicels, from P. racemosum in its umbellate flowers. 4. F. parasiticum. Seem. mss. in Herb. Mus. Brit. — Iledera para- ' silica, Don, Prodv. Fl. Nepal, p. 188; De Cand. Prodr. iv. p. 265. Hedera glauca, Wall. Cat. n. 4921 ; Don, Gen. Syst. iii. p. 394 ; Walp. Eep. ii. p. 432 (v. s. sp.). — Nepal (Hamilton ! in Herb. Mus. Brit. ; Wallich! n. 4921). I have identified Hedera parasitica, Don, with H. glauca. Wall., by means of the authentic specimens, one existing in the British Museum, the other in Burlington House. Hedera acuminata, which Thwaites unites with Hedera parasitica, is at once distinguished generically by its five free styles. Leaflets qnite entire, with the exception of the ultimate one, almost sessile ; umbels globose terminal, simple or com- pound, springing from scaly buds. 5. P. LeschenauHii, Seem. mss. in Herb. Mus. Brit. — Panax LescJie- naultii, De Cand. Prodr. iv. p. 254. Hedera LeschenavUii, Wight et Arn. Prodr. i. p. 877. Panax bijugum, Wall. Cat. n. 4937 ; Don, Gen. Syst. iii. p. 386. Hei/era tri/oliata, Wight et Arn. Prodr. i. p. 377 ; Wight, Icon. Plant, t. 307. Hedera fragrans, Don, PL Nepal, p. 187, non Eoxb- (v. s. sp.).— Nepal (Wallich! n. 4937); Sikhim ON THE GENERA WITH A SINGLE STYLE. 23 (Hooker fil. et Thomson !) ; Nilgherrie Hills (Lesohenault) ; Kumaon (Stvachey et Winterbottom!); Bootan (Griffith ! n. 2065). Wight and Arnott separate the trifoliate from the 5 -foliate form, stating that they had never seen a plant where both kinds of leaves occur. There are no other differences between those two supposed spe- cies ; and, as I have seen specimens with both 3- and 5-foliolate leaves, I have no hesitation in uniting them under the oldest specific name. XIII. Agalm A, Miq. Pedicelli inarticulati. riores ecalyculati, her- maphroditi. Calyx tubo obconico, limbo 5-6- dentato. Petala 5-6, ovato-triangularia, li- bera, 1-nervia, apice incurvula, sestivatione valvata. Stamina 5- 6 ; antherae oblongse. Agalma rugosum {after Micfuet). Stylus 1 ; stigma 5-6-lobatum. Ovarium 5-6-loculare, loculis 1- ovulatis. Drupa oblonga, exsucca, 5-6-angulata, 5-6-pyrena. Al- bumen aequabile. — Arbores inermes Asise tropicse, foliis digitatim foliolatis, foliolis integerrimis v. dentatis, floribus racemosis v. umbel- latis in paniculas dispositis, petalis viridiusculis. — Agalma, Miq. in Bonplandia, 1856, p. 138; Fl. Ned. Ind. i. pars i. p. 752, t. 11 et 13. — Hederce, Paratropice, sp. auct. Miquel wished to confine the genus Agalma to those species which have truly racemose flowers, represented typically by A. rugosum (^Aralia rugosa, Blum.), and it is only in this particular that his Agalma differs from his Paratropia. Singular as is the inflorescence in A. rugosum, simillimum, and racemosum, few botanists would follow him in thinking that character sufficient to found a genus upon. I have therefore sought for better limits, and, adopting Miquel's two Agalmas as the type, added all that agreed generically with them. Most of their congeners had been referred to Paratropia, even by Miquel him- self, where, on account of their long style, tKey were quite misplaced. De Candolle established Paratropia for plants with sessile stigmas, and three out of the four species he referred to it have sessile stigmas, the fourth being a species of Polyscias, and having long styles. Now 34 REVISION OF THE NATUEAL OEDEU HEDERACEjE. Paratropia, restricted to the species witli sessile stigmas, is identical with Heptapleurum of Gsertner ; and as the latter was established as early as 1791, that name, quite as appropriate as Faratropia, claims the right of priority by thirty-nine years. Agalma and Heptapleurum have certain features in common, but they differ in several essential points, viz. : — Agalma. Stylus 1, elongatus. — Arbores plerumque terrestres, flo- ribus viridiusculis. Heptapleurum. Stigmata ovario immersa, punctiformia. — Arbus- culse epiphytse, floribus viridiusculis v. ssepe purpureis v. sanguineis. * Flores racemosi. {Agalma, Miq.) 1. A. rugosum, Miq. in Bonplandia, 1856, p. 138; M. Nederl.- Ind. i. pars i. p. 753, t. 11. — Aralia rugosa, Blum. Bijdr. p. 871. Polyscias rugosa, Keiuw. Herb. Hedera rugosa, De Cand. Prodr. iv. p. 365. Hedera squarrosa, Jungh. in Tijdrsch. Nat. Gesohied. vii. p. 301 ; Walp. Kep. ii. p. 433. Hedera heptapjiylla, Jungh. Itin. — Java, 5-8000 feet above the sea (Horsfield ! in Hgrb. Mus. Brit.; Lobb ! in Herb. Hook.; Junghuhn; Blume). 3. A. simillimum, Miq. in Bonplandia, 1856, p. 138; Fl. Ned. Ind. 1. c. — Hedera simillima, De Cand. Prodr. iv. p. 365. Aralia simillima, Blum. Bijdr. p. 871. Hedera rugosa, Eeinw. Herb. — Java (Reinwardt) . 3. A. raeemosum. Seem. mss. in Herb. Mus. Brit. — Hedera racemosa, Wight Icon. Plant. 1. 1015. Ootacamund (Wight) ; Ceylon (Thwaites ! n. 549, in Mus. Brit. ; Walker ! in Herb. Hook.) ; Nilgherries (Gardner ! Hohenacker !) ** Flores umbellati. 4. A. octopTiyllum, Seem. mss. — Aralia octopkylla, Lour. PI. Co- chinch, (ed. WiUd.) p. 333 ; De Cand. Prodr. iv. p. 358. Faratropia Cantonienais, Hook, et Am. Bot. Beech, n. 189 ; Walp. Eep. ii. p. 433. — Cochinchina (Loureiro !) ; Kakeah Isle (Wright ! n. 101); Canton (Lord Macartney !) ; Macao (David Nelson !) ; Hongkong (Seemann 1 n. 3456. Ilanee! Urquhart ! Wilfcrd ! Hinds! Champion!); Poochoo (Swinhoe!), 5 . A. aromaticum. Seem. — Faratropia aromatica, Miq. in Bonplandia, 1856, p. 139, Fl. Ned, Ind. 1. c. p. 760, t. 13. Aralia aromatica, Blum. Bijdr. p. 871, exolud. var. Hedera aromatica, De Cand. Prodr. iv. p. 365. — Java (Blume! Junghuhn! in Herb. Hook). 6. A. Horsfieldii, Seem. — Faratropia Horsfleldii, Miq. in Bon, ON THE GENERA WITH A SINGLE STYLE. 35 plandia, 1856, p. 139, Fl. Ned. Ind. 1. c. p. 761.— Java (Horsfield! in Mus. Brit.). 7. A. rostratum, Seem. — Hedera rostrata, WigM, Icon. t. 1013-14 ; Walp. Ann. i. p. 359 ; Sikkim (Hooker fil. et Thomson !) ; Nilgherries (Gardner !). 8. A. cBscuUfolium, Seem. — Hedera ascnlifolia, Wall. Cat. n. 4913; G. Don, Gen. Syst. iii. p. 394; Walp. Eep. ii. p. 433; Nepal (Wallich! n. 4913; Strachey et Winterbottom ! n. 3). 9. A. tomentosum, Seem.- — Panax tomentosum, Wall, ex De Cand. Prodr. iv. p. 354. Hedera tomentosa, Ham. in Don PI. Nep. p. 187 ; De Cand. Prodr. iv. p. 364 ; Wall. Cat. n. 4922. Paratropia Wal- licJiiana, C. Kocli, Woohenschrift, 1859, p. 365. — Nepal (Hamilton- Buchanan ! Wallich!, Cat. n. 4932) ; Sikkim (Hooker fil. et Thomson !) ; Bootan (Griffith! n. 3068). 10. A. elatum, Seem. — Hedera elata, Haiu. in Don, PI. Nepal, p. 187; De Cand. Prodr. iv. p. 364. — Nepal (Hamilton-Buchanan! in Mus. Brit.; Wallich! Cat. n. 4914; Strachey et Winterbottom! n. 5). 11. A. lucescens, Seem. — Paratropia lucescens, Miq. in Bonplandia, 1856, Fl. Ned. Ind. l."c. p., 754. Hedera lucescens, De Cand. Prodr. iv. p. 365. Aralia lucescens, Blum. Bijdr. p. 373. — Java(Blume). 13. A. glaucum. Seem. sp. nov. ; foliis digitatim 7-foliolatis, foliolis petiolulatis obovato-oblongis acuminatis integerrimis, basi obtusis v. attenuatis glabris subtus glaucis, iloribus umbellatis, umbellis race- moso-paniculatis, junioribus albido-tomentosis. — Khasia (Hooker fil. et Thomson!). Poliola 6-7-poll. longa, 1-1^ poll, lata ; petioluli ^-1-|- poll, lougi ; vense 10-13. 13. A. Qriffithii, Seem. sp. nov.; foliis digitatim 7-foliolatis glabris, foliolis ovato-oblongis v. oblongis cuspidatis, integerrimis v. obscure denticulatis, basi subcordatis v. obtusis, supra viridibus lucidis, subtus subglaucis, iloribus umbellatis, ramulis pedunculis pedicellis calyci- busque ferrugineo-furfuraceo-tomentosis. — Bootan (Griffith! n. 3064). Petioluli 1 poll, longi ; foliola 4 poll, longa, 3-3^ poll, lata ; vense 7-9. 14. A. redivivum. Seem. n. sp. ; scandens, epiphytum, foliis digita- tim 4-5-foliolatis, foliolis longissime (3-8^ poll, long.) petiolulatis, ellipticis v. oblongo-ellipticis longe acuminatis, remote dentatis, basi 26 REVISION OF THE NATUEAL OEDEE HEDEKACE^. attenuatis, umbellis brevissime pedunculatis, racemosis, pedunculis pe- dicellis calycibusque tomentosis, floribus 7-8-m.eris. — Bangermassing, Borneo (Motley!, n. 1130, in Herb. Hook.). When gathering his specimens, Motley stripped off and left on the ground some leaves, and on re- visiting the spot their petioles were found to have made roots, some of which are iu Hooker's Herbarium. Leaf-stalks 1|- foot long; leaflets, including the petiolules, 1 foot long and from 4-5 inches broad; racemes 1 foot long. XIV. Dendeopanax, Dene. et Planch. Pedicelli inarticu- late riores ecalyculati, polyga- Dendropanax cuneatnin. mo-monoici. Calyx tubo obco- nico, limbo 5-6-dentato. Petala 5-6, ovato-triangularia, l-nervia, sestivatione valvata. Stamina 6-6, antherse oblongse. Stylus 1 ; stigma 5-lobatum. Ovarium 5-loculare, loculis 1-ovulatis. Drupa baccata, subglobosa, obtuse 5-angularis, 5-pyrena. Albumen sequa- bile. — ^Arbores inermes Asise et Araericse tropicae et subtropicse, foliis simplicibus integris coriaceis, umbellis terminalibus, simplicibus v. compositis, floribus pedicellatis, viridiusculis v. albidis. — Dendropa- nax. Dene, et Planch. Eev. Hortic, 1854, p. 107 ; Planch. Hort. Donat. p. 8. Sederce, sp. auct. * Umbellis smpUcilns. 1. B. pendulum, Dene, et Planch. Eev. Hort. 1854, p. 107. — Hedera pendula, Swartz, PL Ind. Occ. p. 513 ; Sw. Icon. t. 9 ; De Cand. Prodr. iv. p. 262 ; Griseb. Pi. West Ind. i. p. 306. — Mountains of Jamaica, Swartz !, in Mus. Biit. ; Wilson ! in Herb. Hook.). 2. D. trilobnm, Seem. — Hedera triloba, Gardn. in Hook. London Journ. iv. p. 105 ; Walp. Eep. v. p. 926. — Porests of Organ Moun- tains, Brazil, 4000 feet elevation (Gardner!, n. 433, in Mus. Brit, et Herb. Hook.). 3. B. nntans, Dene, et Planch. Eev. Hort. 1853, p. 107. — Hedera nutans, Swartz, PI. Ind. Occ. p. 514; De Cand. Prodr. iv. p. 262 ; ON THE GENERA WITH A SINGLE STYLE. 27 Griseb. El. Westlnd. i. p. 306. — Mountains of Jamaica (Swartz I, in Mus. Brit. ; Alexander !, in Herb. Hook.). Dr. Alexander Prior's specimen is a very fine one ; blade of the leaf 3^-3|- inches long, 2-2|- inches broad, leaf-stalk 3-4 inches long. Flowers large for the genus. " A tree 40-50 feet high," Alex- ander. 4. D. Barienense, Seem. n. sp. ; ramis cinereis, foliis longe petiola- tis ovato-oblongis v. oboTato-oblongis, acuminatis, basi acutis, venis costatis 4-5, 2 infimis angulo acuto ortis, umbellis solitariis longe pedunculatis, circ. 30-floris, pedunculis inartioulatis, pedicellis fili- formibus elongatis. — Hedera pendula, Seem. Bot. Herald, p. 132, non Swartz. — By rivulets. Cape Corrientes, Darien (Seemann !, n. 1103, in Mus. Brit, et Herb. Hook.). Petioles . 1-2 inches long. Leaves coriaceous ; blade 4-5 inches long, 2-2^^ inches broad; peduncles 2\ inches long, erect; pedicels very slender, l\-i inches long ; calyx-tubes obconical ; styles ulti- mately divided on the top and recurved. A comparison of my Darien plant with Swartz's authentic specimen of Hedera pendula at once proved their specific differences. The extreme length of the pedicels distinguishes D. Barienense at first sight from all its allies. 5. B.proteum, Benth. Fl. Hongk. p. 136. — Hedera protea, Champ, in Kew Journ. iv. p. 122 ; Hongkong (Champion ! Wright !). 6. B.parviflorum,'Btni\y.'S'l. Hongk. p. 137. — Hedera parvijlora. Champ, in Kew Journ. iv. p. 122. — Hongkong (Champion !). 7. B. Japonicum, Seem. — Hedera Japonica, Jungh. Nov. Gen. et Sp. Plant, p. 25, n. 22 ; Walp. Eep. ii. p. 431. — Japan, not Java as stated by Walpers (Bllime ! in Herb. Benth.). 8. D. ovatum, Seem. — Hedera ovata, Wall. Cat. n. 4911 ; G. Don, Gen. Syst. iii. p. 392 ; Walp. Eep. ii. 431.— Madras (Wallich ! n. 4911). Though the bulk of species belonging to Bendropanax is American, the admission of Hedera protea, parvijlora, ovata, and Japonica, — four Asiatic species, — is justified by their agreeing in every respect with the definition now adopted for the genus. ** Umbellis composiiis. 9. B. Juergenseni, Seem. n. sp. ; glabrum, foliis ovato-oblongis v. ob- longis acuminatis integerrimis, basi attenuatis, venis pinnatis horizonta- liter divergentibus, umbellis terminalibus compositis, ramjs (pedicellis 28 REVISION OF THE NATURAL OEDEK HEDERACEiE. pvimariis) medio articulatis bracteatis, pecUcellis (secundariis) basi bracteolis minutis ferrugineis instractis. — Sievra San Pedro Nolasco, Mexico (Jiirgensen ! n. 729 in Herb. Hook.). Allied to D. arboreum, but all the veins diverge horizontally from the midi-ib, even the lowermost. Leafstalk 2-3 inches long ; largest leaf-blade 7 inches long, 4 inches broad, and having 11-12 veins on each side of the costa. — In D. arboreum the two lowermost veins form an acute angle with the midrib, the same is the case in D. cuneatum. 10. B. Fendleri, Seem. mss. n. sp. ; glabrum, robustum, foliis longe petiolatis ovatis acuminatis integerrimis crasse coriaoeis, basi 3-B- nerviis, supra viridibus lucidis, subtus pallidioiibus opacis, umbellis terminalibus eompositis, pedicellis inarticulatis, floribus . . . , drupa baccata (nigra) 5-angulata 5-pyrena, stylis 5 basi arete connatis apice divaricatis coronata. — Tovar, Venezuela (Fendler ! n. 529). Petiole 3 inches long ; blade 4-5 inches long, 3-3|- inches broad ; drupe f of an inch across. 11. D. arboreum, Dene, et Planch. Eev. Hortic. 1854, p. 107. — Hedera arhorea, Swartz, PI. Ind. Occ. p. 518 ; De Cand. Prodr. iv. p. 262. Aralia arborea, Linn. Am. Ac. v. p. 369 ; Jacq. Hort. Schoenh. t. 51. Hedera alaris, Schlecht. Linnsea, ix. p. 605 ; Walp. Rep. ii. p. 431. Dendropanax alare. Dene, et Planch. 1. c. Sciadophyllum Jacquini, Griseb. Fl. West Ind. i. p. 306. — Jamaica (Swartz! Masson! Wright ! Macfadyen ! Purdie ! March ! AVilson ! Alexander ! Hart- weg ! n. 1544); Island of St. Vincent (Anderson!); Isthmus of Pa- nama (Seemann !, n. 1171; Fendler! ii. 131; Hayes!); Central America (Shakespear !) ; Cuba (Wright ! u. 212) ; Bogota (Triana !) ; Mexico (Jiirgensen ! n. 63, Botteri ! F. Mueller ! Lay and Collie !) 12. B. cuneatum. Dene, et Planch. Eev. Hort. 1854, p. 107. — Hedera cuneata, De Cand. Prodr. iv. p. 263. Aralia umbellata, Pohl, in lit. Bendropanax Sellowianum, Miq. in Ann. Mus. Lugd. Bat. i. p. 27?— Brazil (Gardner! n. 4701; Claussen ! Herb. Mus. Brit.;' Sellow !, n. 1270) ; Eio Uaupe (Spruce ! n. 2860). 13. B. ramiflorum, Seem. — Hedera ramifiora, De Cand. Prodr. iv. p. 262. Aralia ramifiora, Pohl, in lit. Hedera resinosa, Benth. mss. in Plant. Sprucean. — Eio Negro, N. Brazil (Spruce ! nos. 2350, 2337, 2349). 14. D. tomentosum, Seem. ; robustum, ramis crassis tomcntosis mox glabratis, foliis obovato-oblongis acutis mucronatis, basi cuneatis, ON THE GENERA WITH A SINGLE STYLE. 29 supra glabris lucidis, subtus dense ferrugineo-tomentosis, floribus ra- cemosis v. umbellis racemosim paniculatis, pedunculis petiolis ealycibus petalisque ferrugineo-tomentosis, pedicellis brevissimis. — Minas Geraes (Gardner ! n. 4703 ; Claussen! in Mus. Brit.). Petioles i-l inch long. Blade of leaf thick, coriaceous, from 3-5 inches long, 1-|— 3 inches broad. Veins 7-8, on each side of the midrib. 15. B. montanum. Dene, et Planch. 1. c. 16. B. lanceolatum, Dene, et Planch. 1. c. 17. B. obovatum, Dene, et Planch. 1. o. 18. D. oblongum, Dene, et Planch. 1. c. 19. B. densiflorum, Dene, et Planch. 1. c. 20. B. lancifolium. Dene, et Planch. 1. c. 21. B. citrifolium, Dene, et Planch. 1. c. **** Species exclusa. B. Pavonii, Dene, et Planch. 1. c. (Aralia umbellata, Herb. Pav.) = Gilibertia umbellata, Ruiz et Pavon. XV. Hedeka, Linn. Pedicelli inarticulati. Plores ecalyculati, herma- phroditi. Calyx tube ob- conico, limbo 5-dentato. Petala 5, ovato-triangula- ria, 1-nervia, libera, sesti- vatione valvata. Stamina 5 t anthersB oblongse. 0- varium 5-loculare, loculis 1-ovulatis. Stylus 1, elon- gatus; stigma obscure 5- lobatum. Drupa baccata, subglobosa, Isevis, 5-py- YLeiimVL^ViX {partly after Nees von Msenbeck). rena. Albumen ruminatum. Embryo magnus. — Erutices sarmentosi inermes Europse, Asise et Africse iudigense, foliis exstipulatis simpli- cibus palmatim lobatis v. integ-ris, umbellis racemosim v. paniculatim dispositis, floribus viridiusculis, drupis nigris flavis v. albidis. — Hedera, Linn. Gen. n. 395 (ed. Schreb.) ; De Cand. Prodr. iv. p. 261, excl. sp. ; Koch, Synops. Germ. EI. p. 321. 30 REVISION OF THE NATURAL OEDER HEDEKACEiE. Whether there is only one species or several mixed up with the plants which now go in gardens and herbaria under the name oi Hedera Helix, and make up the genus Hedera, as now circumscribed, and whether one or t.\o species are indigenous to the British Islands are still open questions. No botanist has, as yet, been successful in finding good characters for what have been considered as species ; and though all other Sederacea have a limited geographical range, Hedera Helix is sup- posed to be an exception to this rule, and to be spread over three con- tinents, Europe, Asia, and Africa, from the Canary Islands to Japan, and that this circumstance alone sufficiently accounts for the numerous existing varieties. After carefully investigating the subject, and exa- mining every specimen I could lay nay hands on, aided by contribu- tions from . botanical friends, I have arrived at a diiferent conclusion. I can clearly distinguish three distinct species, which, though having each many varieties, do not run into each other, and have each a dis- tinct geographical range. If these different species had to be named anew, I would propose to call them respectively the European, the African, and the Asiatic. The European Ivy is Hedera Helix, Linn. It is not found out of Europe, and may at once be known by its uppermost leaves being . ovate or elliptical, its umbels arranged in simple racemes, and its pedicels and calyx being covered with white stellate hair, the hair having from 6-8, but never more, rays. From time immemorial a variety with white and yellow variegated leaves has been cultivated in gardens; , even PUny mentions it ; indeed it is one of the oldest, if not the oldest, variegated garden plant of which we have any record. The fruit of Hedera Hair of H. Helix. Helix in northern Europe is generally black ; in Germany it occurs occasionally with white ; and in European Turkey, Greece, and Italy with yellow berries. The black-fruited kind has always been considered as the true H. Helix, and the white as a variety of it, which indeed it is ; but the yellow has been made, I think, un- justly into a distinct species, and named H. pddarum by Bertoloni, and some time previously H, chrynocarpa by Walsh. It is the latter plant which played so important a part in ancient Greece and Eome, its leaves supplying the materials for the wreaths with which poets were crowned, and at the festivals in honour of Dionysos all casks, ON THE GENERA WITH A SINGLE STYLE. 31 vesselSj amphoras, etc. were decorated; it was fcustomary even to lie and sit upon ivy branches on those occasions.* It is believed tradi- tionally that the yellow-fruited Ivy came from India with the worship of Bacchus ; and the fact that the Nepal Ivy described by Wallich has yellow fruit is regarded as a proof of the correctness of this tradition. But a close examination of the European yellow-fruited plant shows that it is specifically identical with H. Helve, and specifically different from the Nepal and all other Asiatic specimens. If the worship of Dionysos gradually crept from India to Greece and Eome, and a yellow- fruited Ivy was deemed essential to its proper performance, there was no need of carrying the Asiatic plant into Europe, as an indigenous variety (chri/socarpa=poetarum) occurred at the very threshold ; whilst the Asiatic Ivy, as we shall presently see, is spread from the central high- lands to the most western confines of Asia, — to ancient Colchis. The African Ivy is Reiera Canariensis, Willd. It is found in the Canary Islands, Madeira, and the north of Africa, and may at once be known by its uppermost leaves being cordate, its umbels arranged in panicles, rarely and only in young plants in simple racemes, and its pedicels and calyx being covered with white stellate hair, the hair having from 13-15 rays. To this must be referred what is called in gardens sharp-leaved Irish Ivy. It is a much quicker-growing plant than H. Helim, and on that account frequently planted in gardens. Dr. Moore, of Glasnevin, informs me that it was dis- covered many years ago by a nurseryman named Hair of S. Cana- Hodgins, in the county of Wicklow, Ireland, and the >^ensis. fact that it has lately been found in Portugal, and therefore belongs to the so-called Iberian type of the British flora, would seem a confirma- tion of its being truly indigenous in Ireland. This sharp-leaved Irish Ivy is a form rather than variety of Hedera Canariensis, with more or less deeply-lobed leaves, and must not be confounded with what is called Irish or Scotch Ivy {Hedera Hidemtcav. Scotica) in our gardens. The latter is a large-leaved and quick-growing form of H. ffeliw, more susceptible to cold than the common North European form of that species. The Asiatic Ivy is Hedera ColcUca, 0. Koch. It is not found out For farther particulars, see C. Botticher, ' Baumkultus tier Hellenen ' (Tree- worship of the Hellenes), Berlin, 1856, 8vo, p. 333. D 33 JIEVISION OF THE NATURAL OEDEE HEDERACE^. of Asia, and may be known by its uppermost leaves being elliptical or lanceolate, its umbels arranged in simple racemes, and its pedicels and calyx being covered with yellowish 3-lobed scales, the lobes being opposite each other, and divided into 7-10 segments. Our first knowledge of this plant is derived from Kaempfer, who, two centuries ago, found it in Japan, where it is called " Fo- togi Tsta (i. e. simulacri seu idoli Hederd) ." It was after- wards gathered by Wallich and other collectors in the Hima- laya Mountains, by Fortune in Northern China, by Wright mcokhica. in the Loochoo Islands, and by C. Koch in the trans- Caucasian countries. Wallich was the first to describe _the plant, though he did not venture to separate it from IT. Helix ; in fact, the important character furnished by the scales escaped him. It was not until 1859 that C. Koch, who had seen it wild, named it Hedera Colchiea, and gave a correct diagnosis of it. It is owing to an au- thentic specimen kindly transmitted by him that I am able to iden tify this new species with the Asiatic Ivy. Hedera Colehica is now an inmate of our gardens, it having been found on the Caucasian coast of the Black Sea, by Mr. Eoegner, formerly Curator of the Botanic Gardens of Odessa. Thus it found its- way into our gardens, occumng here and there under the (I believe unpublished) name of Hedera Roegneriana. The only two popular accounts of this plant are given by Wallich and C. Koch. In Nepal it is called Sagooke or Gooke (i. e. the climber), " and is," says "Wallich (Roxb. Fl. Ind. ii. p. 615), one of the most common, as well as the most noble productions .of Nepal, where it grows to a majestic size, and extends over trees and rocks." In Transcaucasia, on the contrary, it is more stunted than the European Ivy, having reached its western geographical limit ; "and," says C.Koch (' Wochenschrift,' 1859, p. 76), "I have never seen it ascend the tops of the numerous Beeches of that country, whilst the common Ivy climbed to the highest branches." Though there are several important peculiarities to distinguish the three species, the most ready way to make sure of them is to look at the character furnished by the hairs and scales. They are largest in H. Helix, where they may be seen distinctly with a comrnqn pocket lens ; but in the two other species it requires a greater magnifying power to make their nature quite intelligible. ON THE GENERA WITH A SINGLE STYLE. 33 Pilis pedicelli et calycis stellato-6-8-radiatis . . . . B. Helix. Pilis pedicelli et calycis 8tellato-13-15-radiati3 . . . H. Canariensis. Squamis pedicelli et calycis bilobis, lobis 7-10-fidis . . H. Colchka. 1. H. Helix, Linn. Sp. 392; Koch, Synops. 11. Germ. p. 321; Sowerby, Engl. Bot. t. 1267; De Cand. Prodr. iv. p. 26], ex parte! —H. communis, S. P. Gray, Natural Arrang. of British Plants (1821), ^ p. 491. — Europe. y' Var. u. melanocarpa. Seem., baccis nigris. Engl. Bot. t. 1267. — Hedera Heliai, var. vulgaris, De Cand. Prodr. 1. c.-^ommon in Northern Europe, and varying with white and yeUow blotched leaves. Or does the white-leaved form belong to the following variety and the yellow-leaved to y. chrysocarpa ? Has any one ever seen them in fruit ? Var. p. leiicocarpa. Seem., baccis albis. — In Germany, rare. -^^-^ Var. y. chrysocarpa. Ten. in Caniel, El. Tosc. p. 300, non De Cand. — Baccis flavis. Hedera chrysocarpa, Walsh in Trans. Hort. Society of ^' London, vi. 42 (1826). H. poetarum, Bertol. PrEelectiones rei Herb. p. 78 (1827). -H.jBoe^icflr, Casp. Bauh. Pin. 305. H. Dionysias, 3. Bauh. Hist. Ic. Hedera chrysooarpos, seu Dionysiacos veterum. Bderte genus chrysocarpon, Plin. Hist. Kissos of the ancient Greeks. — European Turkey, Greece (Heldreich !), and S. and Central Italy. ^ 2. H. Canariensis.WiM. in Berl. Mag. ii.p. 170. t. 5. f. 1 ; Schult. Syst. V. p. 50S.^-JB:. Helix, var. (?) Canariensis, De Cand. 1. c. H. Helix, Lowe, PI. Mad. p. 376, non Linn. — Algiers, Madeira (Lowe !) Portugal (Welwitsch !), Canary Islands (Bourgeau !), and Ireland. ^- 3. H. Cohhica, C. Koch, • Wochenschrift,' 1859, p. 16.— H. Helix, var. Golchica, C. Koch in Linnaea, xvi. p. 365; Eegel, Gartenflora, 1862, tab. 360. H. Helix, Don, El. Nep. p. 187 ; WaUich in Roxb, Fl. Ind. ii. p. 515. H. Helix, var. (?) chrysocarpa, De Cand. Prodr. iv. ' p. 261, exclud. syn. Bauh. et Dalechamp. H. Helix, Thunb. El. Jap. p. 102. "Eotogi Tsta, i.e. simulacri seu idoli Hedera. Hederm arbores communes bacciferge," Kserapf. Araoen. Exot. Ease. v. p. 887. H. rhomhea, Sieb. et Zucc. in Abhand. Bayer. Akad. Math. Ph. CI. vol. *^ iv. p. 202, sine descript. Cissus Wallichima, Tmcz. in Mosc. Bullet., 1838, p. 416.— Japan' (Thunberg ! in Mus. Brit.) ; Loochoo (Wright ! n. 102) ; N. China (Fortune ! n. 60) ; Bootan (Griffith ! n. 2076, in Mus. Brit.) ; Nepal (Hamilton ! in Mus. Brit. ; WaUich 1 Strachey et Winterbottom !) ; Khasia (Hooker fil. et Thomson !) ; Sikkim Hooker fil. et Thomson !) ; Transcaucasia (C. Koch !) ; Black Sea, shores of Caucasia (Roegner). d 2 34 REVISION OF THE NATURAL OEDER HBDBRACE^. Species exclusse : — H. acerifolia, De Cand. =Oreq^fl!?«oa! (?) aoerifolmm. Seem. H. aculeata, Tlam. =Brassmopsis aculeata. Seem." H. acuminata, Wight =Poi^ysaa« acuminata, Seem. H. acuti/oUa, De Cancl.=(?) Dendropanasc. H. eesculifolia, Wall.= ^yaZ»8a cesculifolium. Seem. S. alaris, ^c)ak,tM,.=Bendropanax arboreum, Dene, et Planch. n. angularis, De Cand.= Oreopanax (?) angulare, Seem. H. arhorea, Sw.= I)endropanax arboreum. Dene, et Planch. M. argentata, De Cand.^ Oreopanax argentatum. Dene, et Planch. H. aromatica, De C?ciidL.= Agalma aromaticum, Seem. H. Australiana, '¥. yLTii&sx-=Kissodendron Australianum, Seem. H. avicennicBfolia, De Cs,nA.= Oreopanax avicenniafolium, D. et PL H. capitata. Smith = Oreopanax capitatum. Dene, et Planch. H. catalpeefolia, De CaLT\A.=Oreopanaso catalpafolium, D. et PL H. cheirophylla, De Cand.== Oreopanax (?) CheiropTiyllum, Seem. H. corymbosa, C]iois.=:IIgdnocarpus corgmbosus. Seem. H. crasdnervia, De Candi.=0?-eopanax c?'«SM>2em«m,Dcne.et Planch. S. (?) Cumanenm, De GdXiA.^Oreopanax (?) Oumanense, D. et PL H. cuneata, De Qs,n&..=^I)endropanax cuneatum. Dene, et Planch. H. discolor, De Cani,= Oreopanax discolor, Dene, et Planch. a. disperma, De CaRi.= Macropanax oreophilum, Miq. H. Echinops, &. 'Y)on^=Oreopanax Mohinops, Dene, et Planch. H. eiaia, 'Rsm.= Agalma efafew,. Seem. H . ferruginea, De Cs.n&..= Sciadophyllum ferrugineum, D. et PL //. ferruginea. Wall , = Trevesia . H.floribunda, De Cs,rLA.= Oreopanax floribundum. Dene, et Planch. H.floribunda, Wall.^Brassaiopsis Jloribunda, Seem. H.fragrans, D. Don=Pe«te^a»flsa; LeschenauUii, Seem. H.fragrans, S^v.^^Heteropanax fragrans. Seem. H. GaudicJiaudii, A. Gra,y=Cknrodendron. H. glauca, Yfall.^Pentapanaxparasiticum, Seem. IT. glomerulata, De Oand.=S?"a«mojB«j* speciosa. Dene, et Planch. ^. Hainla, Jiam. =Brassaiopsis Hainla, Seem. Jff^. heptapltylla, JTingh. =:Jgalma rugosum, Miq. if. heterophylla, 'Wall.^Heptapleurum. keterophyllum, Seem. i/". hypoglauca, 'Rance=Fitis Oantoniends, Seem. - iT. Jackiana, G. Don= Artkrop^yllum. ON THE GENEKA WITH A SINGLE STYLE.- 35 H. Japonica, Jvmgh. =Dmdropanax Japonieum, Seem. H. jatrophcefolia, De Cand.= Oreopanaxjatrophcefolium, T)c. et PI. H. latifolia, Wight et Arn.=Gren. nov. (?). ZT. LeschenauUii, Wight et AYa.=Pentapamtx LescJienauUii, Seem. H. lucescens, De Cs.iiA.=^Agalma htcesceus. Seem. If. multiflora, De CwLA.=Oreopanax capitatum. Dene, et Planch. JS. Muiisiana, De Cuni. ^Oreopanam Mutisiana, Dene, et Planch. H. nutans, ^w .^Dendropanax nutans, Dene, et Planch. H. ohovata, y^\^i^= Heptapleurum stellatum, Gsertn. H. oMusiloha, De Cand.= Oreoj»««aa! o6\Qi. yi. p. 176.- — Actinofhyllum pedicel- lutum, Euiz et Pavon, PL Per. iii. p. 73, t. 308. — Mountains about Muiia, Peru (Pavon ! in Herb. Mus. Brit.). 4. S. micranthum. Planch, et Linden, Araliac. p, 20. — Ocaria, New Granada (Schlim, n. 1149). This is in Herb. Hook, under Schlim, n. 699, if I have correctly determined the species. 5. S. villosum, Planch, et Linden, Araliac. p. 20. — New Granada (Schlim, n. 44). Corolla unknown. fi. S. Goudotii, Planch, et Linden, Araliac. p. 21. — Bogota, and province of Pamplona, New Granada (Goudot ! Schlim! n. 3 et 5). 7. 8. oxyanthum. Planch, et Linden, Araliac. p. 21. — New Granada (Triana, n. 372). 8. 8. Quinduense, De Cand. Prodr. iv. p. 261. — Aralia Quinduensis, H. B. et K, Nov. Gen. v. p. 8, t. 417, f. 1 ; Planch, et Linden, Araliac. p. 23. — Nomen vemac. " Pata de gallina." New Granada, Andes of Quiudiu (Humboldt et Bonpland ; Triana !), Antioquia (Purdie ! Jervise !) . 9. 8. decapJtyllum, Seem. (sp. nov.) ; glabrum; foliolis 9-10, obovato-oblongis retusis basi acutis, supra lucidis ; floribus umbellatis ; drupis 5-angulatis; stylis 5 recurvis persistentibus coronatis. — Panax decapliyllum, Sagot, Herb. — Prench Guiana (Sagot ! n. 916). — Leaflets 3-4 inches long, \\-'i inches broad, and their surface so remarkably shining that this species is at once distinguished by it. The specimens are in fi'uit only. ON THE GENERA. WITH SEVERAL DISTINCT STYLES. 51 10. S. Abyssinicum; Steud. Nomencl. Bot. p. 537 ; Miq. Ann. Lugd. Bat. i. p. 26. — Astropanax Abymnicum, Seem, in Journ. of Bot. 1865, p. 177. Aralia Abyssinica, Hochst. mss. in Schiinp. PI. Abyss. ed. i. n. 283 ; liich. Tent. Fl. Abyss, i. 336 ; Wnlp. Ann. ii. p. 724.— Adoa, Abyssinia (Elochstetter !). 11. S. elalum. Seem. — Astropanax elatiim, %&e,m. \. c. Paratropia data. Hook. fil. in Journ. Linn. Soc. vii. p. 196. — Cameroon Moun- tains, 7500 feet above the sea (Mann !). — This species is very close to S. Abynsinicuin, but appears to be sufficiently distinct, 12. S. Barteri, Seem. — A&tropanax Barteri, Seem. Journ. of Bot. 1865, p. 177. — Glabrum ; foliolis 5 longe peliolatis ovato- v. obovato- oblongis acuminatis basi rotundatis integerrimis coriaceis supra nitidis ; umbellis racemosis"7-12-floris, bracteis ovatis acuminatis ; drupis sub- globosis 8-pyrenis.— Sugar Loaf Mountains, Niger (Barter! n. 2027) " A small tree," Barter. 13. S. Baikiei, Seem. — Adropanaw Baikiei, Seem, in Journ. of Bot. 1865, p. 177. — Grlabrum ; foliolis 5 (?) longe petiolatis elliptico-ob- longis acuminatis in petiolum att;nuatis integerrimis submembrana- ceis ; umbellis racemosis 4-5-floris, bracteis lanceolatis ; drupis ovatis 4-6-pyrenis (albis) acute angulatis; — Niger (Barter !). — " Twenty feet high, fruit white when ripe," Barter. I have named this species in commemoration of the late Dr. Baikie, the indefatigable explorer of the Niger, to whose expedition the late Mr. Barter was attached as bota- nist. The leaflets are larger than these of the preceding species, less coriaceous, and diiferent in shape. ** Flores capitati. 14. S. Manni, Seem. — Astropanax Manni, Seem, in Journ. of Bot 1865, p. 178. Paratropia Manni, Hook. fil. Journ. Linn. Soc. vi. p. 10.— Pernando Po. 5000 feet above the sea (Mann !). 15. S. conicum, Poir. Diet. vi. p. 746, — ActinopJiyllum conicum, Euiz et Pav. PI. Per. iii. p. 74, t. 309.— Peru (Pavon ! in Herb. Mus. Brit., Mathews !). 16. S. acuminatum, Poir! Diet. vi. p. 746. — Actinophyllum acumina- tum, Euiz et Pav. Fl. Peruv. iii. p. 74, t. 310. — Peru (Pavon I in Herb. Mus. Brit., Lechler ! n. 2592). 17. 8. angulatum, Poir. Diet. vi. p. 745, exclud. syn. Browne. — Actinophyllum angulatum, Ruiz et Pavon, PL Peruv. iii. p. 73, t. 307, non H. B. et K.— Peru (Pavon ! in Herb. Mus. Brit.). 53 REVISION OF THE NATURAL ORDER HEDE.EACEiE. 18. 8. Rnmboldtianum, Dene, et Planch, in Planch, et Linden, Araliac. p. 33. — Actinophyllum angulatum, H. B. et K. Nov. Gen. v. p. 93, non Euiz et Pavon? — New Granada, between Pansitara and AlmagTier (Humboldt and Bonpland). 19. S. pentandrum, Poir. Diet. vi. p. 747. — ActinopJiyllum pentan- drum, Uuiz et Pavon, PL Peruv. iii. p. 75, t. 311. — Muna, Peru (Pavon 1 in Herb. Mus. Brit.). 30. 8. Triance, Planch, et Linden, Araliac. p. 33. — New Granada (Triana! Jervise !). I have seen the corolla of this plant, unknown to Planchon and Linden. It is hemispherical. 31. 8. heterotrichum, Planch, et Linden, Araliac. p. 33. — New Granada, between Bogota and Pusagasuga (Goudot ! n. 1 B.). 33. S. Mathewsi, Seem. (sp. nov.) ; ramis foliisque glabris ; foliolis 8-10 oblongis v. obovato-oblongia, obtusis v. abrupte acuminatis, basi obtusis v. acutiusculis integerrimis coriaceis, supra nitidis subaveniis, subtus venis prominulis ; capitulis (Pisi maj. magnitudine) racemoso- paniculatis, rachis ramisque villoso-puberulis, pedunculis angulatis ; corolla hemisphaerica puberula, apice depresso ; staminibus 5, antheris oblongis; stylis 4-5; drupis obcomco-4-5 -angulatis, glabris. — Cha- chapoyas, Peru (Mathews! in Herb. Hook, et Btli.). Leaflets 3y-4 inches long, IJ inches broad. 33. 8. ferrugineum, Dene, et Planch, in Eev. Hort. 1854, p. 107. — Aralia ferruginea, H. B. et K. Nov. Gen. v. p. 7. Hedera ferruginea, De Cand. Prod. iv. p. 364. — ^Between Saraguro and Oiia (Humboldt and Bonpland). 34. 8. Sprucei, sp. nov. Seem. ; foliolis 9 longe petiolulatis ovali- oblongis acuminatis in petiolum attenuatis glabris, supra atroviridibus, subtus pallidioribus venis primariis subpurpurascentibus ; rachi pe- duuculisque tomento stellato deciduo vestitis ; floribus sessilibus ; ca- pitulis (foem. fructif.) parvis ; corolla ign. ; stylis 4 recurvis. — ^Tarapoto, Peru (Spruce! n. 4550). I have named this species in honour of its discoverer, that enterpris- ing South American explorer, Dr. Spruce. Species indescripta. 35. 8. caudalum, Dene, et Planch, in Rev. Hort. 1. c. (Herb. Dombey). ON THE GENERA. WITH SEVERAL DISTINCT STYLES. 53 26. S. Gayanum, Dene, et Planch. 1. c. 37. S. ochroleucum, I>cne. et Planch. 1. c. Species excMsa. S. nvene, TLorth. =:Heplapleurum avene. Seem. S. capitatum, Gnsth.^Oreopanax eapitatum. Dene, et Planch. S. digitatum, G. J)on= Oarolmia insiffniSjUort. 8. dwaricatJ4m, B\ume= Heptapleurum divaricatum, Seem. S. ellipticum, ^\\ime:= Heptapleurum ellipticum, Seem. S.farinosum,,'S\..=^Heptapleurumtomentosum, Hassk. var. S. kumile, Tilume^ Heptapleurtim parasiticum, Seem. S. Jacquinii, Griseh.^ Dendropanax arboreum. Dene, et Planch. S. longifolium, 'Bl.=IIeptapleurum longifolium, Seem. S. lucidum, Blnme = HeptapUurum longifolium, Seem. var. S. macrostacTiyum, ^ih..=Brassaia macrostacJiya, Seem. S. palmatum, Blume=2Veuma' Sundaica, Miq. S. parasiticum, 'Sium^^Heptapleurum parasiticum, Seem. S. pulchellum, Hort. =: Tupidatit/ius calyptratus, Hook. f. et Th. ;S'. racemiferum, Miq. = Gen. nov. S. rotundifolium, Hort. = (P) Heptapleurum sp. S. rubiginosum, Planeh. etLind, = Gen. nov. S. scandens, 'S>\Mxa&= Heptapleurum scandeus, Seem. S. suhavene, ^la.me=: Heptapleurum Coronce-sylva, Seem. S. tomentosum, 3lame^= Heptaplewriim tomentosum, Hassk. S. verticillatum, S^anoghe= Heptapleurum verticillatum, Seem. XXIV. PsEUDOPANAX, C. Koch, ' Wochenschrift,' 1859, p. 336. Pedicelli articulati, apice incrassati. Flores ecalyculati, polygami. Calyx tube obconioo, limbo obsolete denticulato. Petala 5, ovato- triangularia, 1-nervia, sestivatione valvata. Stamina 5 ; antherse ovato- oblongee. Ovarium 5-loculare, loculis 1-ovulatis. Styli 5, basi co- adunati, apice liberi. Brupa baccata, globosa, 5-pyrena. Albumen Eequabile. Frutices- v. arbusculse Novse-Zelandise, foliis alternis ex- stipulatis, simplicibus vel digitato-3-5-foliolatis, foliolis crassis coria- ceis oblongis vel linearibus, sinuato-dentatis, umbellis racemosis v. , paniculatis, terminalibus. — Xylophylla, Banks et Sol. mss. Aralia, Cussonia et Panada sp. Auct. This genus is closely allied to Cheirodendrou, Nutt.,the relationship having been suggested by De CandoUe ; but in Pseudopanax the calyx 54 REVISION OP THR NATURAL OBDKR HEDERACE^. is not calyculate, though the pedicel is swollen at the articulation, and the stigmas are not seated on a stylopodium but on five long styles, more or less free above the base. C. Koch thought that Panax simplex, anomalum, Gaudichaudii, pla- typliyllum, arborum, Uneare, Edgerleyi, and Cohnsoi might possibly belong to Pseudopanax. But if P. crassifoUum be regarded as the type of the genus, those species, not being pentacarpous or having distinct styles, are inadmissible. 1. P. crassifoUum, 0. Koch, Wochenschrift, 18S9, p. 366. — Aralia crassifoUa, Banks et Sol. mss. ; Parkinson's Drawings of New Zealand Plants, tab. 101, 103 (ined.); Hook. Icon. Plant, t. 583, 584; Hook, fil. Fl. New Zeal. i. p. 96. Panax coriaceum, Kegel, Gartenflora, 1859, p. 45. P. lon'gissimum. Hook. f. Handbook Fl. New Zealand, i. p. 102. — New Zealand (Banks and Solander; Forsterl in Herb. Mus. Brit.). This is a very variable plant. Parkinson, who accompanied Captain Cook in one of his voyages, has left drawings of two varieties which he gathered in New Zealand, the one having crimson, the other purple petioles. In recent years. Continental nurserymen have raised as many as twenty varieties, chiefly differing in the division, shape, and colour of the leaves. One of the oldest inmates of our gardens has been described as a separate species by llegel (1859) under the name of Panax coriaceum, and by Hooker f. (1868) under that of P. longissi- mum. It has a simple stem and a dark-green bark striped with brown ; aU the leaves are quite simple. Eegel enumerates two varieties of P. coriaceum, viz. u. latifolinm (leaves |— f of an inch broad, the older ones with rather larger teeth, Aralia lalifolia, Hort.,) and fi. angusti- folium {Aralia integrifolia, Hort., leaves \ of an inch broad, with very small teeth). Panax pentadactylon, Dene, et Planch. Hort. Donat. p. 10 {Aralia pentaphylla et A. quinquevulnera, Hort.), and P. tridac- tylon. Dene, et Planch. 1. c. {Aralia triphylla and A. trifoliata, Hort.), and Aralia heteromorpha are also garden varieties of this variable species. a. P. Lessoni, G. Koch, Wochenschrift, 1859, p. 3&6.—Pdnax (I) Lessoni, De Cand. Prod. iv. p. 253. Cussonia Lessoni, A. Uich. Fl. N. Zel. p. 285, t. 32. Aralia trifolia, Banks et Sol. mss.; Parkinson's Drawings N.Zeal. Plants, t.l03, 104 (ined.). Aralia Lessoni, Hook. fil. Fl. N. Zel. i. p. 96. — New Zealand, in woods. (Banks and Solander ! in Mus. Brit.) ON THE GENERA WITH SEVEHAL DISTINCT STYLES. 55 XXV. Oligoscias, Seem. (gen. iiov.) — Pedicelli aiiiculati. Plores calyculati, hermaphroditi. Calyx tubo turbinato, limbo 5-deiitato- Petala 5, ovato-triangularia, 1-nei-via, libera. Stamina 5, filamentis brevibus, antheris oblongis. Ovarium inferum, 4-5-lociilare, loculis 1-ovulatis. Styli 4-5, erecti, omnino liberi, stigmatibus 2-fidis. Drupa baccata, subglobosa, 4-5-pyrena. Albumen ruminatum. — Frutex. Madagascariensis, inermis, foliis exstipulatis pinnatis 2-3-jugis cum imparl, petiolis supra canaliculatis, foliolis lateralibus sessilibus, terminali petiolulato, infimis subrotundatis substipulseformibus, supre- mis ovatis basi aoutis v. acuminatis, 3-5-setaceo-deiitatis; umbellis 5-8-floris, simplicibus v. compositis; pedicellis filiformibus (6-8 lin. long.) ; floribus drupisque albis. Species unica : — 1. 0. Madagascariensis, Seem. (sp. nov.) — Madagascar, Betroun, Tamatave and Antananarivo, on clay hills, 2000-3000 feet above the sea. (Meller ! Lyall ! n. 232, and others.) A transverse section of the albumen presents almost the same cross- shaped figure as that of Sciadopanax, though not quite so regular as in that genus. XXVI. PoLYSCiAS, Forst. Pedicelli articnlati. Florea calycu- lati, hermaphroditi. Calyx tubo turbinato v. hemisphserico, limbo minute 5-8-dentato. Petala 5-8, vulgo 5, ovato-triangularia, 1-nervia, libera, sestivatione valvata. Stamina 5-8, vulgo 5, antheris oblongis. Ovarium 5-8, vulgo 5-loculare, loculis 1-ovulatis. Styli 5-8, erecti, an- gulati, demum recurvi. Brupa baccata, subglobosa, 5-angulata, 5-8- pyrena. Albumen . . — Frutices v. arbusculse Asise et Polynesise tropicee, inermes, glabrae, caule simplici v. parum ramoso, foliis exsti- pulatis imparipinnatis, foliolis ovato-oblongis v. subrotundatis subin- tegerrimis v. crenatis, floribus umbellatis v. subcapitatis, in paniculas amplas dispositis. — Polyscias, Forst. Gen. p. 63. t. 33 (1766) ; De Gaud. Prod. vi. p. 257. Bupteron, Miq! in Bonplandia, 1856, p. 139. Hederce, Panacis, et Nothopanacis sp. Auct. In the specimen of P. pinnata figured and described by Porster, there was an excess of petals and stamens, and this, together with the imperfect way in which the whole genus has been characterized, pro- bably accounts for Miquel's failing to recognize it, establishing the genus Mupteron (absolutely identical' with Porster's Polyscias), and, mistaking Porster's species for a Nothopanax, describing it as Notho- panax tricocMeatwm in his ' Supplement of the Dutch East Indian 56 REVISION OF THE NATURAL ORDER HEDERACEiE. Flora.' Miquel, on that occasion, proposed to enlarge the generic cha- racter of Nothopanax, so that it includes 5-inerous species, and those who follow him would have to call all Nothopanacia species of Poly- scias, Polyscias being the older name. But by combining dimerous and pentamerous ovaries in the same genus, we should once more re- turn to the chaos from which the Sederaceae are just emerging. De CandoUe regarded Polyscias nodosa as a species of Paratropia, but it differs from the other species he associated it with by its pinnate leaves and long recurved styles. The above generic character has been worked out from Forster's original specimens and drawings pre- served at the British Museum. * Flores pedicellati. 1. P. pinnata, Forst. Gen. p. 64. t. 32; Prod. n. 518; Icon, (ined.) t. 298; De Gaud. Prod. i?. p. 257. — Polyscias umbellata, Spreng. ex Steud. Nom. Bot. p. 546. Panax (?) Hayneanum, Wall. Cat. n. 4927, et. G. Don, Gen. Syst. iii. p. 383; Walp. Eep. ii. p. 429. Panax Forsteri, Dene, et Planch. Rev. Hort. 1854, p. 105. Notho- panax tricocJdeatum, Miq. Fl. Ned. Ind. Suppl. i. p. 340. Aralia Polyscias, Spreng. in Erschein. — Tana (Forster ! Anderson ! in Mus. Brit.), Sumatra (Wallich ! n. 4927, et in Herb. Soc. Linn. Lond., Teijsmann ! Diepenhorst ! in Herb. Hook.). The leaflets being very apt to di'op off in drying, have been described by G. Don as simple leaves. 2. P. Papuana, Seem. — Gastonia Papnana, Miq. Ann. Lugd. Bat. i. 5. New Guinea (Zippelius). Has generally 8 styles, but some- times 12 by excess. 3. P. acuminata. Seem. mss. — Hedera acuminata, Wight, Icon. t. 1062. Bupteron acuminatum, Miq. in Bonplandia, 1856, p. 139 ; Fl. Ned. Ind. i. pars i. p. 762. — Nilgherries (Wight), Ceylon (Thwaites! Ceylon Plants, n. 4.). Wight thinks Hedera Jackiana, Don, may possibly be identical with Polyscias acuminata, but that species belongs to Arthrophyllum ; and Thwaites refers Hedera parasitica, Don, to P. acuminata as a synonym, but from Don's authentic specimen it proves to be a species oi Pentapanax (JP. parasiticum. Seem.). ** Flares aeasilea. 4. P. nodosa. Seem. mss. — Aralia nodosa, Blum. Bijdr. p. 873. ON THE GENUS OBEOPANAX. 57 Paratropia nodosa, De Cand. Prod. iv. p. 365. Hedera nodosa, Hassk. Tijdr. Nat. Gesch. x. p. 131. Aralia wrnbraculifera, Eoxb. El. Ind. ii. 108. Papaja aylvestris, Rumph. Amb. i. p. 149. t. 53, fig. 1. Eupte- ron nodosum, Miq. in Bonplandia, 1856, p. 139. — Java and Moluccas (Blume), Philippine Islands (Cuming ! n. 504 et 1053 in Mus. Brit.). According to Eoxburgh about twelve feet, according to Blume thirty feet high. Rumphius's figure is referred by De CandoUe to Bergera Kcenigii, a plant which has a very different inflorescence, whilst Eoxburgh and Miquel referred it to this species. By some misprint, however, the latter author quoted plate 13, which represents a Palm. VII. On the Genus Oreopanax. I should not have been able to give a complete enumeration of all the species composing this large genus and SciadopJiyllum if Dr.Planchon and M.J. Liuden had not obligingly sent me the proof-sheets of a work on the plants of New Granada and Venezuela, written by thera, which con- tain descriptions of the Ai'aliacea of that region, and of which they al- lowed me to make a free use. 1 have not reprinted the descriptions, but merely referred to them under the title " Planch, et Lind. Jraliacem." MM. Decaisne and Planchon have also kindly furnished me with infor- mation relative to several of the undescribed species- enumerated by them in the ' Eevue Horticole.' Botanists who meet with any species of Oreopanax in a wild state or cultivated in tropical gardens should endeavour to observe whether the flowers are really dicBcious. Many AraliacecB are polygamous, but some genera seem to be always herma- phrodite. Observations on these points are very much needed. XXVII. Oreopanax, Dene, et Planch. Eevue Horticol. 1854,p. 108 ; Planch, et Lind. ^rafoac. p. 5 {ined.).—¥\oxts diclines, dioici (?). Calyx limbo brevi subintegro v. obscure sinuato. Petala 5, rarissime 4, libera, sestivatione valvata. Masc. : Stamina petalorum numero ; fila- menta subulata, petalis longiora. Styli imperfecti 1-5, liberi v. in- ferno concreti. Fcem. : Stamina sterilia 4-5, fertilibus mascvdis bre- viora. StyK 3-6, liberi v. plus minus concreti. Drupa baccata, SEepius 4-5-6-pyrena, nunc abortu 1-2-3-pyrena. Albumen rumina- tum. — Arbores v. frutices Americse tropicse, plerique Andicoli ; foli'. simplicibus integris v. varie divisis, stipulis in unam intra-axillarem peliolo plus minus adnatum parvam connatis; floribus sessilibus, singulis bractea v. bracteolisque duabus stipitatis ; cupitulis panicu- latis ; ovario semper ^sbvo.—Aralice et Hederm sp. aiict. 58 HEVISION OF THE NAT0EAL ORDER HEDERACES: 1. 0. capitatum, Dene, et Planch, in Rev. Hort. 1854. p. 108. — Aralia capitata, Jacq. Amer. p. 89. t. 61. Hedera capitata. Smith, Icon. Pict. p. et tab. 4. H. multiflora, De Cand. Prod. iv. p. 263. SciadopJiyllum capitatum, Griseb. Fl. West Ind. p. 306. — The most widely diffused species of Oreopanax. Veraguas, on the volcano of Chiriqui (Seemann ! n. 1648); Organ Mountains, Brazil (Gardner! n. 5736); Peru (Dombey 1 Mathevps !) ; Ca,rib Islands (Jacquin! De Ponthieu I) ; Jamaica (Alex. Prior ! W. Wright ! Wilson 1 F. Masseii ! Marsh ! ) ; Central America (E. Shakespear !) ; Belize (Marsh ! )' ; Venezuela (Fendler 1 n. 1330, 536 ; Punk and Schlim ! n. 91 ; Ernst ! n. 571) ; Mexico (F. Miiller !) ; Martinique (Sieber! n. 94) ; N. Granada (Goudot! n. 5); Ecuador (Spruce! n. 5535). Called in Belize " Three- foot Jaflc," because the branches are seldom more than three feet long. (Marsh.) In Venezuela it is termed, according to Ernst, " Candelero." 3. 0. septemnervium, Dene, et Planch. 1. c. — Aralia septemnervia, H. B. et K. Nov. Gen. v. p. 3. Uedera septemnervia, De Cand. Prod, iv. p. 363. — New Granada (H. B. etK.). 3. 0. (?) catalpafolium. Dene, et Planch. 1. c. {ined.). — Aralia ca- talpmfolia, WiUd. Herb, ex Ro3m. et Schult. Syst. vi. p. 697. Hedera catabpmfolia, De Cand. Prod. iii. p. 363. — New Granada (Triana, n. 537 ; Humboldt and Bonpland). 4. 0. Mutisianum, Dene, et Planch. 1. c. — Aralia Mutisiana, H. B. et K. Nov. Gen. v. p. 1. A. helerophylla, Mutis, non Willd. Hedera Mutisiana, De Cand. Prod. iv. p. 263. — New Granada (Mutis, Hum- boldt and Bonpland, Linden 1 n. 1259; Goudot! n. 1; Hartweg! n. 935; Triana! 189; Purdie !). 5. 0. sessilijiorum. Dene, et Planch, in Eev. Hort. 1. c. — Hedera ses- siliflora, Benth. PI. Hartweg. p. 146. — Ecuador, near Loja (Hartweg! 816). 6. 0. avicenniaifolium. Dene, et Planch, in Eev. Hort. 1. c. — Aralia avieenniafolia, H. B. et K. Nov. Gen. v. p. 3. t. 3. Hedera avicen- niafolia, De Cand. Prod. iv. p. 263. Aralia trachonanthifolia, Willd. in Eo3m. et Schult. Syst. vi. p. 698. — Ecuador (Humboldt and Bon- pland; Spruce! n. 5999). 7. 0. Echinops, Dene, et Planch, in Eev. Hort. 1. c. — Hedera Echi- nops, G. Don, Gen; Syst. iii. p. 394. Aralia Echinops, Schlecht., Linnaea, v. p. 174. — Mexico. 8. 0. macrocephalum. Dene, et Planch, in Eev. Hort. 1. c. sine de- ON THE GENUS OllEOPANAX. 69 script. ; ramis petiolis rachibus pedunculisque ferrugineo-tomentosis deinum glabratis ; foliis ovato- v. subvotundato-oblongis acuminatis dentatis v. grosse repando-dentatis, 3-nerviis coriaceis, supra glabiis, subtus dense flavido-tomentosis ; capitulis masculis globosis in racemos elongates dispositis ; bracteis ovatis acuminatis ; petalis extus villosi- usoulis ; capitulis fructiferis oblongis in racemos abbreviates dispositis ; baccis globosis stylis 6 liberis coronatis. — Province of Larecaja, Bolivia, in forests (Mandon ! n. 571; Kelly! in Herb. Bth. — ) Petioles 3-4 inches long ; blade of leaf 5-6 inches long, 3-4 inches broad ; female flower-heads as large as a good-sized walnut. The plant here described must be 0. macrocephalum, which M. Decaisne informs me is founded upon a specimen gathered by Weddell in the same locality as that collected by Mandon. 9. 0. LecMeri, Seem. (sp. nov.) ; robustum ; ramis petiolis rachibus pedunculisque ferrugineo-tomefitosis demum glabratis ; foliis ovato- oblongis acuminatis, basi subcuneatis, integerrimis v. hinc inde den tatis, 5-7-nerviis, supra glabris, subtus dense ferrugineo-tomentosis capitulis fructiferis longe pedunculatis pauois ; stylis 5-7 ; cset. ign. — Gachapata,, Peru (Lechler ! n. 2386). — Leaves about a foot long, 4-5 inches broad. Fruiting ilower-heads as large as a cherry. 10. 0. cordatum. Planch, et Linden, Araliaceae, p. 7. ined. — New Granada (Goudot ! n. 6; Linden, n. 962; Purdie!). — Leaves some- times 3-lobed in Purdie's specimens. 11. 0. Pavoni, Seem. (sp. nov.); ramis petiolis rachibus pedunculisque dense fusco-tomentosis ; foliis longe petiolatis ovatis v. subcordato- ovatis acuminatis 5-nerviis crasse coriaceis, supra stellato-puberulis demum glabratis, subtus dense fusco-tomentosis ; paniculis masculis amplis ; capitulis globosis (Pisi maj. magnitudine) ; petalis 4 ovatis extus stellato-pubescentibus ; fl. foem. ign. — Peru CPavon ! in Mus. Brit.). — Petioles 4-5 inches long; blade of leaf 4-5 inches long, 3-3 1 inches broad. 12. 0. Eouadoriense, Seem. (sp. nov.) ; foliis longe petiolatis ovatis 3- fidis, lobis divaricatis ovato-oblongis acuminatis irregulariter et gi'osse dentatis, basi obtusis, 3-nerviis crasse coriaceis, supra glabris, subtus cinereo-tomentosis ; rachibus pedunculisqlie tomentosis ; capitulis masc. ovatis v. oblongis ; petalis 4 ovatis acuminatis, extus tomentosis ; foem. ign. — Western side of the Andes, above Guayaquil, Ecuador (Pavon ! in Mus. Brit.).— Petiole 7 inches long, blade of leaf 7 inches long. 60 REVISION OF THE NATURAL OKDEE HEDERACEiE. 13. 0. Macleani, Seem. ; rarais petiolis rachibus pedunculisque dense villoso-toraentosis ; foliis longe petiolatis ovatis apice B-lobis, lobis ovato-triangulaiibus acuminatis integerrimis v. hinc inde dentatis, 8- nerviis, supra glabris, subtus dense villoso-tomentosis ; capitulis raasc. globosis Pisi maj. magnitudine; petiolis extus villosis ; foem. ign. — Peru (Maclean! in Herb. Hook.). 14. 0. Sprucei, Seem. (sp. nov.) ; vamis petiolis rachibusque albido- villoso-tomentosis ; foliis palmatim 3-5-fidis, lobis oblongis acuminatis dentatis, supra glabris, subtus albido-tomentosis ; capitulis masculis in paniculas amplas dispositis, Pisi magnitudine ; petalis 5 ovatis acuminatis, extus albido-tomentosis, demum glabratis; fl. foem. ign. — ^Nomen vernac. Ecuadoriense : Puma-maqui («. e. pes tigridis) teste cl. Spmce. — In stony places at the foot of Mount Tunguragira, Ecuador (Spruce! n. 5113).— "A tree 20-30 feet high, slightly branched; flowers white," Spruce. Allied to 0. argentatum. 15.0. Mathewdi, Seem. (sp. nov.) ; rarais petiolis rachibusque ferni- gineo-tomentosis ; foliis palmatim 3-B-fidis, supra glabris, subtus sparse steUato-pubescentibus, deraum glabratis ; fl. masc. in paniculas amplas dispositis ; petalis extus glabratis ; fl. foem. ign. — Chachapoyas, Peru (Mathews I in Herb. Bth.). 16. 0. Boliviense, Seem. (sp. nov.) ; ramis petiolis rachibus peduncu- lisque ferrugineo-floccoso-villosis mox glabratis ; foliis palmatim 3-5- fidis, lobis spinuloso-dentatis 6-9-nerviis, demum utrinque glabris ; fl. ign. ; baccis glabris (nigris) stylis 5 coronatis. — Andes of Bolivia, prov. of Larecaja, in forests (Mandon ! n. 569). 17. 0. discolor, Done, et Planch. 1. c. — Aralia diavohr, H. B. et K. Nov. Gen. v. p. 4. Hedera discolor, De Cand. Prod. iv. p. 263. — New Grranada (Humboldt et Bonpland; Triana, n. 538). Peru (Lechler ! n. 1974). 18. 0. argentatum. Dene, et Planch. I.e. — Aralia argentata, H. B. et K. Nov. Gen. v. p. 4. t. 114. A. Jieterojphylla, Willd. ; Rcem. et Schult. Syst. vi. p. 698, fide Kunth. Hedera argentata, De Cand. Prod. iv. p. 263. — New Granada (Humboldt et Bonpland). 19. 0. crassinervium. Dene, et Planch, i.e. — Aralia crassinervia, H. B. et K. Nov. Gen. v. p. 5. Hedera crassinervia, De Cand. Prod, iv. p. 263. — New Granada (Humb. et Bonpl.). 20. 0. lancifolium. Planch, et Linden, Araliaceae, p. 8. — Venezuela (Linden ! n. 1433, in Herb. Mus. Brit, et Hook.). ON THE GENUS OEEOPANAX. 61 21. 0. fraternum, Dene, et Planch. Rev. Hort. I.e.; Planch, et l.ind. Araliacese, p. 9.— Venezuela (Funk et Schlim, u. 1530). 22. 0. Horguetero, Dene, et Planch. Eev. Hort. p. 108 ; Planch, et Lind. Araliae. p. 9.— Nomen vernae. Horquetero.— Venezuela (Funk et Selilira! n. 1529). 23. 0. hypargyreum, Dene, et Planch, in Eev. Hort. 1. c. ; Planch, et Linden, Araliae. p. 10.— New Granada (Punk et Sehliin,n. 1333). What Deeaisne and Plauehon have named 0. hypoleucum is, as Dr. Plauchon tells me, this species, 34. 0. Triana,I>cne. et Planch, in Eev. Hort. I.e.; Planch, et Linden, Araliae. p. 10.— Nomen vernae. Higueron. — New Granada (Triana! n. 375, Jervise ! in Herb. Hook.). 25. 0. chrysoleucum, Planch, et Linden, Araliae. p. 11. — Venezuela (Funk et Schlim). 26. 0. Gecropia, Dene, et Planch, in Eev. Hort. 1. c. ; Planch, et Linden, Araliae. p. 11, — New Granada (Schlim, n. 113). 27. 0. myriocarpum, Dene, et Planeh. in Eev. Hort. 1. e. ; Planch, et Lindl. Araliae. p. 12.— Ibague, New Granada (Goudot ! n. 2, Holtbn !). 38. 0. floribundum. Dene, et Planch, in Eev. Hort. 1. c. — Aralia floribunda, H. B. et K. Nov. Gen. v. p. 6, t. 416. A. Humboldtiana et incisa, Willd. in Eoem. et Schult. vi. p. 699, fide Kunth. Hedera floribunda, De Cand. Prodr. iv. p. 264. — New Granada (Humboldt et Bonpland, Triana! n. 2761, 2703). 39. 0. Rumboldtianum, Dene, et Planeh. in Eev. Hort. 1854, I.e. — 0. Bonplandianum, Planch, et Linden, Araliae. p. 12. Aralia fer- ruginea. Linden, Cat. Hort. non H. B. et K. — Nomen vernae, fide Ernst, Yagrumo macho. — Silla de Caracas and Valle de Cobre, Venezuela (Fendler! n. 531, 532, 533, Birschel! Punk et Schlim ! n. 1524et92, Ernst! n. 1118.) According to information kindly supplied by M. Deeaisne, the unde- scribed 0. Himloldtianum is identical with 0. Bonplandianum of Plan- chon and Linden. It is Bonpland's n. 646. 30. 0. coriaeeum, Dene, et Planeh. in Eev. Hort. 1. e. ; Planch, et Linden, Araliae. p. 13. — New Granada (Triana! n. 190). Pliinehon and Linden quote Triana's n. 130, which js probably a misprint. 31. 0. Ooannense, Dene, et Planeh. in Eev. Hort. 1. e. ; Planch, et 63 BETISION OF THE NATURAL ORDEE HEDERACEjE. Linden, Avaliac. p. 13. — Ocaiia, New Granada (Schlim! n. 491, in Herb. Hook.). 32. O.flabellatum, Done, et Planch, in Kev. Hort. i.e. ; Planch., et Linden, Araliac. p. 14. — New Granada (Schlim ! n. 836, in Mus. Brit, et Herb. Hook.). 33. 0. bracJiystachyum, Dene, et Planch, in Kev. Hort. 1. c. — 0. hracJiybotryum, Planch, et Linden, Araliac. p. 15. — New Granada (Triana, n. 737). 34. 0. 5;'MK«e«)», Dene, et Planch, in Eev. Hort. I.e. ; Planch, et Linden, Araliac. p. 5. — Antioquia, New Granada (Triana, n. 376, Jervise !). 35. 0. sclerophyllum. Done, et Planch, in Uev. Hort. 1. c. ; Planch, et Linden, Araliac. p. 15. — New Granada (Schlim! n. 114, in Herb. Hook.). 36. 0. jatrof\ to Hederacea, a view now generally adopted. 1. H. aculeata; foliis peltatis utrinque inermibus subtus dense stellato-tomentosis ; umbellis subsessilibus basi pluri-bracteatis ; calyce hispido; fructu paleaceo-setoso. — H. aculeata, Bl. Bijdr. p. 885; Brown et Bennett, 1. c. p. 133. t. 35. H. peltata, Benth. et Hook. Gen. i. p. 937— Java (Horsfield ! in Mus. Brit.). 3. H. horrida; foliis palmatis utrinque aculeatis, subtus ad costas nervisque pilis crispis sparsis instructis ; umbellis pedunculatis, ebrac- teatis ; calyce fructuque inermi glabro. — Horsfieldia horrida, Seem. mss. Panax horridum. Smith in Bees' Cycl. n. 10 ; De Cand. Prodr. iv. p. 353 ; Hook. Fl. Bor. Am. i. p, 373. t. 98 ; Torr. et Gray, M. N. Amer. i. p. 648. Aralia erinacea, Hook, in Brewst. Edinb. Journ. 1837, p. 64; De Cand. Prodr. iv. p. 359. Eohinopanax horridum. Dene, et Planch. Rev. Hort. 1854, p. 105. — North-west America (Menzies! Capt. Portlockl in Mus. Brit), Wahlanat (Nuttall! in Mus. Brit.); Japan (C. Wright! in Herb. Kew.). There is a good illustration of the mode of growth and habit of this plant (so great an impediment to travellers in the woods of north- west America) in my translation of Kitlitz's ' Twenty-four Views of the Islands and Vegetation of the Pacific,' plate ii. fig. 7-|- and 3/3i. XXIII. On the Genus Acanthopanax. XLI. Acanthopanax, Seem. Pedicelli inarticulati. Flores eca- lyculati, polygami. Calyx minute 5-dentatus. Petala 5, 1-nervia, 86 REVISION OF THE NATURAL ORDER HEDEKACEiE. libera, valvata. Stamina petalorum numero iisque alterna ; antherae ovatEe V. oblongse, Styli 3, basi connati, apice divaricati. Discus convexus v. medio couicus. Ovarium 2-loculare, loculis 1-ovulatis. Fmctus didymo-compressus, 2-locularis. Albumen sequabUe. — Ar- busculsB V. frutices aculeati, Indise or., Chiuse, et Japoniae, foliis sim- plicibus V. digitatim 3-B-foliolatis, foliolis serratis, umbellis globosis solitariis v. paniculatis. — Fanax subg. Acanthopanax, Dene, et Planeli. Eev. Hort. 1854, p. 105. Ealopanax, Miq. in Ann. Mus. Lugd. Bat. i. p. 16. Pleetronia, Lour. Fl. Cochin. 1. A. ricinifolium. Seem. — Fanax ricinifolium, Sieb. et Zucc. in Abhandl. Baier. Akad. iv. 3. p. 199. Kalopanax ricinifolium, Miq. in Ann. Mus. Bot. Lugd. Bat. i. p. 16. Brassaiopsis ricinifolia, Seem. Journ. of Bot. vol. ii. p. 291. — Japan (Siebold ! in Herb. Benth. ; Old- ham ! Maximowicz !), N. China (Fortune ! in Herb. Mus. Brit.). 2. A. aculeatum. Seem. Panax aculeatum, Ait. Kew. iii. p. 448 ; De Cand. Prod. iv. p. 252 ; Jacq. Coll. iv. p. 175 ; Icon. Ear. t. 634. — Zanthoxylon trifoliatum, Linn. Spec. 1455; Lam. Diet. ii. p. 40. Pleetronia Chinensis, Lour. Fl. Cochinch. i. p. 201. Panax Lourei- rianum, De Cand. 1. c. p. 252. Aralia trifoliata, Meyen, Eeise, ii. p. 332, nomen, teste C. Koch, Wochenschrift, 1859, p. 866. — South- ern China, about Canton and Macao (Seemann ! n. 2457; Millett! in Herb. Hook. ; Sir G. Staunton ! Eobertson ! Bradley ; Lord Ma- cartney ! in Mus. Brit.), Assam plains (Jenkins ! in Herb. Hook.), Khassia (Griffith ! in Herb. Hook.) ; East Indies (Wallich !, n. 4926), Assam (Griffith!, n. 787 in Mus. Brit.), Amoy (Herb. Bth.). 3. A. spinosum, Miq. Ann. Lugd.-Bat. i. p. 10. — Panax Linn. fil. Suppl. 441 ; Lam. Diet. ii. p. 715. Aralia ;^ Thunb. Fl. Jap. 128 ; De Cand. Prodr. iv. p. 259.— Japan (Thunberg ! in Mus. Brit. ; Wright ! in Herb. Hook.). Thunberg's specimen at the British Museum has only 2, not 5, styles. 4. A.sepium,^esia.; arbuscula; ramis patentibus aculeatis, aculeis recurvis ; foliis 3-5-foliolatis, foliolis ellipticis acuminatis in petiolum attenuatis dentatis glabris, petiolis aculeatis ; umbellis globosis termi- nalibus, solitariis v. paniculatis ; pedicellis elongatis ; calyce ecalyculato 5-dentato; petalis 5, 1-nerviis, liberis; stylis 2, apice divaricatis, basi connatis ; fructu didymo-compresso, 3-loculari. — " Small, duU green tree or bush, with dense patent branches, growing in hedges." Hook, fil. in Sched. Herb. Kew.— Khasia Hills, E. Indies (J. D. Hooker !). ON THE GENUS TETRAPANAX. 87 5. A. divaricatum, Seem. — Panax divaricaium, Sieb. et Zucc. Ab- hand. Baier. Akad. iy. 2. p. 198 ; Walp. Ann. i. p. 981. Kalopanax divaricatum, Miq. Ann. Lugd.-Bat. i. p. 17. — Japan (Siebold ! in Herb. Benth. ; Maximowicz !). 6. A. aeisilijlorum, Seem. — Panax sesdlijlorum, Eupr. et Maxim. H. Amur. p. 131; Begel, Gartenflora, 1862, t. 369. — Amur (Maxi- mowicz), Manchuria (Wilford !), Northern China (Fortune? Kscher). Stem and petioles armed. 7. A. Sumatranum, Seem. — Kalopanax Sumatranum, Miq. 1. c. — Sumatra. 8. A. resectum. Seem. — Kalopanaoo resectum, Miq. 1. c. — Patria ? 9. A. innovam, Seem. — Panax innovans,Sieb. et Zucc. I. c. — Japan (Maximowicz !). XXIV. On the Genus Tetkapanax. XLII. Tetuapanax, C. Koch,Wochenschr. f. Gartn. und Pflanzenk. 1859, p. 371 ; Ed. Otto et Sender in Hamburg. Gartenz. 1862, p. 61. Pedicelli inarticulati. Flores eoalyculati, hermaphroditi. Calycis margo obsoletus. Petala 4 v. 5, ovata, acuta, sestivatione valvata. Stamina 4 v. 5 ; iilamenta elongata ; antherae elliptico-ovatse. Styli 2, erecti, demum recurvi. Ovarium 2-loculare, loculis 1-ovulatis. Drupa baccata. — Frutex Chinensis, inermis, stoloniferus, caule erecto 5-9-ped. striato annulate intus copiose albissimo-medulloso ; foliis terminalibus longe petiolatis amplis palmatim lobatis, lobis omnibus acutis serratis, subtus (prsecipue junioribus) totis stellato-subferru- gineo-tomentosis, stipulis 2 magnis subulatis ; umbellis in paniculas ter- minales amplas dispositis ; floribus albido-flavis. — Didi/mopanaa:, subg. Tetrapanax, C. Koch, Wochenschrift, n. 10. p. 71. Aralice sp., Hook. Prof. C. Koch was the first to perceive that the plant described by Hooker as Aralia papyrifera had nothing to do with Aralia, and there- fore referred it as a subgenus to Bidymopanax, — subsequently sepa- rating it as a distinct genus {Tetrapanax), to which he thought P. in- novans and P. ricinifolium might possibly belong. But the two latter both out turn to be species of Acanthopanax, a genus differing in its spinose habit and dry fruit from Tetrapanax, but which is by fjir more intimately related to it than Fatsia, to which Hook. fil. and Bentham (Gen. Plant, p. 939) have referred both Aralia papyrifera and A. erinacea {=HorsJieldia), Ed. Otto has already pointed out (Ham- burg. Gartenz. 1. c.) that in Aralia papyrifera pentamerous flowers pre- 88 REVISION OF THE NATURAL OKDER HEDEKACE^>. dominate over the tetramerous ones, and that the name Tetrapanass is on that account open to objection ; but if we were to change every in- appropriate name, our botanical nomenclature would have to undergo considerable alterations, and I therefore shall retain that of C. Koch. The only species of Tetrapanax at present known is the following : — 1. T. papyriferum, C. Koch, 1. c. — Aralia (?) papyrifera, Hook, in Kew Journ. iv. p. 53. t. 1 et 3. A. papyrifera, Hook. Bot. Mag. t. 4897; PI. des Ser. viii. t. 806, 807: xii. t. 1201 ; G. Bennett, Gatherings of a Naturalist, t. 6; ejusd. Wanderings, vol. ii. p. 77, cum icone (1834) ; ejusd. in Seem. Journ. of Bot. 1864, p. 309. " Rice-paper Plant " of English Gardens. — Island of Formosa (Old- ham ! in Herb. Hook.), China (Bowring!). Interesting particulars about the habit of this plant are given by Dr. Bennett, who, it appears, was the first to publish a figure of the Eice-paper Plant. It throws up numerous suckers, the suckers flower- ing the second year after they have been removed from the mother plant, and every time after flowering form two new branches. XXV. On THE Genus Asteotbicha. XLIII. Asteotbicha, De Cand. Prodr. iv. 74, et Mem. Omb. 29. t. 5, 6 ; Benth. et J. Hook. Gen. Plant, i. p. 937; Benth. El. Austr. iii. p. 379. — Pedicelli articulati. Floras ecalyculati, hermaphroditi. Ca- lycis tubus ovatus ; limbus minimus, vix 5-dentatus. Petala 5, ovalia, subacuta, sestivatione valvata. Stamina 5 ; antherse oblongse. Discus subplanus, margine libero undulato. Ovarium S-loculare. Styli 2, dis- tincti, a basi filiformes ; stigmata terminalia. Fructus ovatus, a latere compressus exalatusque v. transverse subteres ad commisuram longitu- dinaliter alatus, exocarpio membranaceo v. subcarnoso ; pyrenae a latere compressae, prsesertim ad commisuram induratse ibidemque utrinque sulcatae v.inloculos spurios vacuosproduotcB. Semen oblongum. Al- bumen Eequabile. — Frutices Australienses, plus minus stellato-tomen- tosi V. lanati, foliis alternis petiolatis indivisis subtus tomenlosis ; stipulis nuUis ; umbellis paniculatis ; bracteis parvis setaceis v. nullis. A genus so closely allied to NotJiopanax, that it can only be separated by artificial characters ; so that the latter, bearing the more recent name, may have to be merged into it. In 1863 I transferred it (Journ. of Bot. i. p. 280) from Umbellifera, where up to that time it had been placed, to Seder aeem, — a view since adopted by various authors. AKALIACE*. 89 1. A. pterocarpa, Benth. Fl. Austr. iii. p. 379. — Queensland, at Ktzroy Island (Walter Hill !). 2. A.Jloccdsa, De Cand. Mem. Ombell. 30. t. 5 ; Prodr. iv. p. 75 ; Benth. PI. Austr. iii. p. 879. — Bolax floccipes, Sieb. PI. Exs. n. 258. — Queensland and New South Wales, Australia (E. Brown ! SieLer ! n. 258, A. Cunningham!). Var. a. subpeltata, Benth. I.e. Var. p. angustifolia, Benth. 1. c. Var. y. incana, Benth. 1. o. — A. latifoUa, Benth. in Hiigel, Enum. 55. 3. A. longifolia, Benth. in Hiigel, Enum. 55 ; PI. Austr. iii. p. 380. — Queensland and New South Wales (A. Cunningham ! P. Mueller ! E. Brown! M'Arthur!). 4. A. ledifolia, De Cand. Memb. Omb. 30. t. 6; Prodr. iv. p. 74; Benth. PI. Austr. iii. p. 380. — A. hoveoides, A. Cunn. ; Benth. in Hiigel, Enum. 55. A. linearis, A. Cxmn. ; Benth. I.e. y. asperifolia, P. Muell. ; Klatt in Linnsea, xxix. p. 709. Bolax ledifolins, Sieb. Plant. Exs. n. 25. — New South Wales and "Victoria (Sieber! n. 257, A. Cunningham ! P. Mueller !). Abaliace^. The following genera belong to the AraliaceiB proper, distinguished by an imbricate (often quincuncial) sestiyation of the coroUa. To them should be added my genus Pentapanax, which does not have a valvate coroUa, as stated in Journ. of Bot. ii. p. 294 ; whilst, on the other hand, Mackinlaya, P. Muell., should be referred to Umbelliferce proper. The genera belonging to this suborder form two natural groups : — I. AkaliEjE. — Ovarium 5- -, abortu 3-, nunquam 2-loculare. 1. Aralia, Linn. 2. Siilboearpa, Dene, et Planch. 3. Pentapanax, Seem. 4. Sciadodendron^ Grisb. II. PanacBjB. — Ovarium 2-, per excessum 3-, nunquam 4-co-locu- lare. 5. Panax, Linn. 6. Myodocarpus, A. Brongn. et Gris. 7. Belarbrea, Vieill. 90 EEVISION OF THE NATUEAL ORDER HEDERACEjE. XXVI. On the Genus Aealia. XLIV. Aralia, Linn. Gen. n. 386, fexcl. sp. ; Benth. et Hook. Gen. i. p. 936.- — Pedioelli articulati. Mores saepe polygamo-monoici. Calycis margo prominulus, tnincatus, repandus v. brevissime 5-dentatus. Pe- tala 5, ovata, obtusa v. brevissime inflexo-acurainata, marginibus plus minus imbricatis. Stamina 5 ; antherse oblongse v. rarius ovatse, rectse. Discus subplanus v. rarius conicus, margine libero. Ovarium 3-5-loculare; styli nunc basi ereoti v. breviter connati, superne de- mum recurvi, nunc a basi recurvi v. summo apice infiexi ; stigmata terminalia. Pructus 3-5-angulatus, exocarpio carnoso ; pyrense 3-5 orbiculatse, ovatse v. oblongse, compressse, orustacese v. durae. Semen compressum, albumine aequabili. — Herbse perennes v. frutices, glabri pubescentes setosi v. aculeati. Polia alterna, digitata pinnata v. ternato-pinnatim pinnatimve decomposita, foliolis serrulatis. Stipulse a basi petioli pai-um prominentes. UmbellulEe solitariae racemosse paniculatse v. terminales, rarius in umbellam compositam dispositas. Bracteee parvse. — Dimorphanthus, Miq. Com. Phytogr. 95. t. 12. This genus is here restricted to the pentecarpous, by abortion tri- carpous, species ; the truly dicarpous ones, having a distinct habit, are referred by me to Panax. Species Ghinenses : — 1. J. Gkinensis, Linn. Sp. 393 ; De Cand. Prod. vol. iv. p. 359, excl. syn. Blume; Hance in Seem. Joum. of Bot. 1866, p. 172. — Leea spinosa, Spreng. — China (Lord Macartney! in Herb. Mus.). 2. A. Planchoniana, Hance in Seem. Journ. of Bot. 1866, p. 172. — Ilha Verde, near Macao, S. China (Hance). 3. A. Becaisneana, Hance in Seem. Journ. of Bot. 1866, p. 172. — Island of Formosa. 4. A. MandscJiurica, Seem. — Dimorphanthus Manchuricus, Eupr. et Maximow. PI. Amer. p. 133. — On the Lower Amur (Maximowicz ! in Herb. Hook.). Species Japonica : — 5. A. data, Seem. — Dimorphanthus elatus, Miq. Com. Phytogr, 95. t. 12 ; Walp. Eep. vol. ii. p. 430. — Aralia grandis, Miq. Herb. Japan. 6. A. canescens, Sieb. et Zucc. Abhand. Math. Phys. Kl. Baier. Akad. vol. ii. p. 223 ; Walp. Ann. vol. i. p. 982. — Japan (Biirger ! Oldham !). ON THE GENUS ARALIA. 91 7. A. edulis, Sieb. et Zuoc. PI. Jap. i. 57. t. 25.- edulis, Miq. Com. Phytogr. p. 96 ; Walp. Eep. ii. p. 431. Aralia cordata, Thunb. Tl. Jap. p. 127 ? A. racemosa, var. Sachalinensis, Eegel, Gartenflora, 1864, t. 432 ?— Japan. Species Indicts : 8. A. Gachemirica, Dene, in Jacquem. Voy. iv. p. 72. t. 82 ; Walp. Eep. ii. p. 430.—^. macrophylla, Lindl. Bot. Eeg. (New Series) xvii., Plant. Misc. p. 73. n. 72. Panax tripinnatum, Wall. Cat. n. 4934. P. decompositum, Wall. Cat. n. 4935 ; De Cand. Prod. iv. p. 255. — Nepal (Wallichl Cat. n. 4934 et 4935). 9. A.foliolosa, Seem. — Panax CO foUolosum, Wall. Cat. n. 4928. — Sikhim (Hooker f. et Thomson !) ; Bootan (Griffith ! n. 2074 in Mus. Brit.) ; Silhet (Wallich ! n. 2928). 10. A. Thomsonii, Seem. (n. sp.). — Khasia mountains (Hook. f. et Thomson ! Araliacea, n. 42) ; Assam Plains (Jenkins !). 11. A. armata, Seem. — Panax armatum, Wall. Cat. n. 49, 83 ; G. Don, Gen. Syst. iii. p. 886 ; Walp. Eep. ii. p. 429.— Tavoy (Wallich ! n. 49, 37) ; Khasia and Sikhim (Hook. f. et Thomson !). 12. ^. Finlaysoniana, Seem. — Panax Finlaysonianum, Wall. Cat. n. 49, 36 ; G. Don, Gen. Syst. iii. p. 429 ; Walp. Eep. ii. p. 492.— India (WaUich ! n. 4937.). 13. A. cissifoUa, Griffith, mss. in Herb. Hook, — Panax scandens, Edgw. mss. in Herb. Hook.— Bootan (Griffith!), Kumaon (Strachey and Winterbottom !). Has the habit of Meutkerococcws. Leaves 5-nately digitate. Species Boreali- America : 14. A. nudicaulis, Linn. Sp. p. 393, non Blume ; Eaf. Med. Bot. i. t. 8 ; Torr. et Gray, PI. North Amer. i. p. 646. Nomen vernac. "WUd Sarsaparilla."- — Canada to the mountainous portions pf the Southern United States (Nuttall ! Douglas ! Anderson ! Kertland ! in Herb. Mus. Brit.). 15. A. racemosa, Linn. Sp. p. 393 ; Schk. Hand. t. 86; Toit. et Gray, Fl. North Amer. i. p. 646. Nomen vernac. " Spikenard." — Canada to mountains of Georgia and Eocky Mountains (Nuttall! Kertland !). 16. A. Mspida, Michx. Pi. Am. Sept. i. p. 185 ; Vent. Hort. Cels. t. 41 ; Sims, Bot. Mag. t. 1041 ; Lodd. Bot. Cat. t, 1306 ; Torr. et Gray, PI. North Amer. i, p. 647. — A. Muhlenbergiana, Schult. Syst. 92 KEVISTON OF THE NATURAL ORDEE HEDEEACE^. vi. p. 704. Nomina vernac. " Wild Elder " et " Bristly Sarsaparilla." — Lakes Winipeg and Superior (Eichardson !) ; Sachakawan (Richard- son !); Nova Scotia (Alex. Anderson !) ; Massachusetts (Nuttall !). 17. A. humilis, Cav. Icon. iv. p. 7. t. 313. — Mexico. 18. A. pubescens, De Cand. Hort. Monsp. 1813, p. 80; Prod. iv. p. 258. — Mexico. 19. A. spinosa, Linn. Sp. p. 393. — A. spinosa,\ax. 7, Torr. et Gray, Fl. North Amer. i. p. 647. Nomina vernac. "Prickly Ash," "An- gelica Tree," "Hercules Club."— United States (Nuttall! Pursh ! Gouan ! Drummond ! Short !) 30. A. Leroana, C. Koch, Wochenschrift, 1864, p. 369. — A. spinosa, var. ^, Torr. et Gray, PI. North Amer. i. p. 647. A. Japonica, Hort. Germ, non Auct.— North America. Professor K. Koch, 1. c, thus distinguishes A. spinosa and A. Le- roana. A. spinosa ; spinosissima ; pe- A. Leroana ; spinosa ; petioli tioli glabri ; foliola oblongo-lanceo- puberuli ; foliola supra aspera aut lata, cuspidata, glabra, subtus deuique glabriuscula, acuta aut glaucescentia ; panicula peduncu- acuminata, subtus pubescentia lata, elongata ; rami alterni, denuo glaucescentia ; panicula sessilia, ramosi, ramulis plerumque ad contracta, ramis elongatis denuo apicem solum umbelliferis ; flores ramosis, ramulis ubique umbelU- majores. feris ; flores minores. Species Javanica et PMlippinenses : 21. A. Javanica, Miq. in Bonplandia, 1859, p. 137. — Aralia CM- nensis, Blume, Bijdr. p. 870, non Linn. — Java (Horsfield ! in Mus. Brit. ; Junghuhn ! in Herb. Hook.) 23. A. ht/poleuca, Presl, Epim. p. 350 ; Walp. Ann. ii. p. 734.— Philippine Islands (Cuming ! n. 930 et 793 in Mus. Brit.). 33. A. montana, Blume, Bijdr. 870 ; Miq. PI. Ned. Ind. 1. c. p. 750. — Aralia bipinnaia, Eeinw. Herb. — Java (Horsfield ! in Mus. Brit.). 34. A. dasypJiylla, Miq. in Bonplandia, 1856, p. 138; PI. Ned. Ind. 1. c. p. 751. — Java (Junghuhn !). 35. A.ferox,M.\<\. in Bonplandia, 1856, p. 137; PI. Ned. Ind. 1. c. p. 750. — Java. Species exohiscB. A. Abyssinica, Hochit.^Sciadophyllum Abyssinicum, Steud. A. acerifolia, Vlilld. =Oreopanaa; acerifolium, Seem. ON THE GENUS ARALIA. 93 A. aculeata, 'B.am.= Brassaiopsis aculeata. Seem. A. aci/tifolia, Willd.=(?) Dendropanax. A. angularis, '^iM. = Oreopanax angulare. Seem. A. arborea, Linn. = Dendropanax arboreum, Dene, et Planch. A. arborea, Avrnh. — Gil^ertia Brasiliensis, Seem. A. argentata, H. B,=Oreopamx argentalum, Dene, et Planch. A. aromatica, Blume=^yaZ/Ma aromaticum. Seem. A. avicennimfolia, H. B.= Oreopanax avicennwfolium. Dene, el PI. A. calycttlata, Zoll. et Mor. =Macropanax oreophilum, Miq. A. capitata, Jacq^.^Oreopanax capitatum. Dene, et Planch. A. capitulata, Jungh. et Vries.=(?) Osmoa^ylon. A. catalptBfolia, l^\M.--=Oreopanax catalpepfolium, Dene, et Planch. A. cheirophylla, Spr.^ Or eopanax clmrophyUum, Dcue. et Planch. A. Chinensis, Lour.=(Rumph. Amb. iv. 44.) A. cocMeata, liam.=Nbthopanax cochleatiim, Miq. A. crassifolia, So\.^=Pseudopanax crassifoliiim, 0. Koch. A. crassinervia, H. 'B.=^Oreopa)iax crassinervium, Dene, et Planch. A. Cumanensis, H. 'Q.--=Oreopanax Cumanense, Dene, et Planch. A. digitata, WiUd.^Oreopanax Xalapense, Dene, et Planch. A. digitata, 'Rox\).=:HeptapleurMm venulosum. Seem. A. discolor, H. 'B.=Oreopanax discolor. Dene, et Planch. A. disperma, 'B\nme=Macropanax oreophilum, Miq. A. dubia, Spr.^Trevesia palmata, Vis. A. Echinops, Cham.= Oreopanax EcMnops, Dene, et Planch. A. erinacea, ^oo\.= Horsfieldia Jiorrida, Seem. A.farinosa, 'Ddi\e=Nothopanax farinosum, Seem. A. ferruginea, H. '&.=8ciadophyllum ferrugineum, Dene, et Planch. A. ferruginea, limAen^^ Oreopanax ITumboldtiaaum, Dcue. et Planch. A. floribunda, H. 'B>.=^Oreopanax floribundum, Dene, et Planch. A. fragrans, Don (Steud. !) =Heteropanax fragrans, Seem, et Pentapanax LeschenauUii, Seem. A. Jieptaphylla, 'V^i\ld.=SciadophyUumferrugineum, Dene, et Planch. A. heterophylla, 'M.ntis== Oreopanax Mutisianum, Dene et Plancb. A.heterophylla, 'Wi].\d.==Oreopanax acerifolium. Seem. A. Humboldtiana, Roem. el ^c\m\i.=Oreopariax floribundum, Dene. et Planch. A. Japonica, Thunb.=^a\\3me,=Heptapleurum pergamaceum, Hassk. A. pinnata, Ji.ociist.=^Nothopanax farinosum, Seem. A. platanifolia, H.'B.=Oreopanax platanifolium. Dene, et Planch. A. polaris, Homhx. =8tilboearpa polaris. Dene, et Planch. A. Polyscias, Spr.=Polyscias umbellata, Porst. A. polygama, Bks. et Sol.= Schefflera digitata, Porst. A. Quinduensis, H. 'B.=Sciadopkyllum Quinduense, DC. A. quinquefolia, A. Gray=JPa!«air quinquefolium, Linn. A. ramiflora, Yo\i\=Bendropanax ramiflorum, Seem. A. Reinviardtiana, ^ievA.= Trevesia Sundaica, Miq. A. reticulata, H. B.^Oreopanax Humholdtianum, Dene, et Planch. A. rigida, B\ume= Heptapleurum rigidum. Seem. ON THE GENUS STILBOCAKPA. 95 A. rugosa, B\vLvat=Agalma rugosum, Miq. A. salicifolia, Vent.^? A. scandens, 'Foix.=Brassaiopsis Hainla, Seem. A. Schefflera, Spr.= Schpfflera digitata, Porst. A. Sciadopkyllum, %w.=8ciadop]iyllum Brownei, DC. A. Sieboldii, Hort.= (?) Fatsia Japonica, Dene, et Planch. A. septemnervia, H. B.=Oreopanax septemnervium, Done, et Planch. A. simillima, 'Blam6=Agalma simillimum, Miq. A. tarchonanthifolia, '^'iM.=Oreopanax avicenma/olium, Dene et Planch. A. trifolia, 'Sks.=Pseudopamx crassifolium, C. Koch. A. trifolia, A Gvay=Panaii; trifolium, Linn. A. trifoUata, Meyen^Acantkopanax aculeatum, Seem. A. trigyna, GiM(i.. = Ckeirodendron Gaiidichaudii, Seem. A. tripliylla, \^o\x.^Fanax trifolium, Linn. A. Turhaeensis, E.B.^Oreopanaas Turiacense, Dene, et Planch. A. wmhellata, 'Pohl=Deiidropanax cuneattm, Dene, et Planeh. A. umbellata, Va.v.= Giliderta umbellata, Euiz et Pav. A. umhellifera, 'ha,nx.= Osmoxylon Amhoinense, Miq. A. umbraculifera, Hoxh.^Polyscias nodosa, Seem. A. Valdiviense, Gay^Ckeirodendron Valdiviense, Seem. A. Vitiensis, A. Gvsy= Schefflera Vitiensis, Seem. A. Xalapensis, H..B.=Oreopanax Xalapense, Dene, et Planch. XXVII. On the Genus Stxlbocarpa. XLV. Stilbocaepa, Dene, et Planeh. Eev. Hort. 1854, p. 105. — Pedicelli articulati. Kores eealyculati, polygami. Calycis tubus ob- ovatus ; limbus truncatus. Petala 5, obovato-oblonga, obtusa, sestiva- tione imbricata. Stamina 5 ; antherae oblongae. Styli S-4, divergentes, recurvijOmnino liberi. Stigmata punctiformia. Ovarium 3-4-loculare. Drupa depresso-sphserica, suberosa, 3-4-pyrena. Albumen sequabile. — Herba perennis, regionibus subantarcticis et Nova Zelandia inhabi- tans, inermis, tota setis mollibus laxis obsita, foetida ; foliis (maximis) stipulatis longe petiolatis orbiculari-reniformibus basi profunde cor- datis marginibus multilobatis ; umbellulis compositis ; involucris folia- ceis ; floi'ibus densis ; drapis atris nitidis, grana piperis magnitudine. — Hook. f. Fl. N. Zeal. i. p. 95. jlrafe'« sp., Homb. et Jacq. Species yb REVISION OF THE NATURAL ORDER HEDERACE^. 1. S. polaris, Dene, et Plancli. 1. c. ; Hook. f. Fl. N. Zeal. i. p. 95 ; A. Gray, Bot. Wilkes, p. 716. — Aralia polaris, Homb. et Jacq. Voy- age au Pole Sud, t. 3, sine descript. ; Hook. f. PI. Ant. p. 19 ; Hand- book Fl. N. Zeal. i. p. 100. Hook. Icoa. Plant, t. 744.— Geogr. DiSTR. Southern Island of New Zealand (Lyall 1) ; Lord Auckland's Islands (J. D. Hooker ! Dr. Holmes !) ; Campbell's Islands (J. D. Hooker!), covering large tracts. This genus, on account of the quincuncial aestivation of its coroUa, belongs to Araliacem. XXVIII. On the Genus Sciadodendron. CPl.*.te VII.) I first met with this tree in 1846, at Panama, where it is known by the name of Jobo de lagarto, and commonly used for making fences, — poles stuck in the ground taking root readily, and growing very fast. On the 23nd of March, 1847, I sent a collection of Panama woods to the Kew Museum, and n. 33 of that collection were specimens of this tree. But though I looked carefully and frequently for flowers, I could never find any. Mr. Sutton Hayes also records amongst his notes that he had known the tree for years, without being so fortunate as to find any flowers on it. My friend Dr. Duchassaing was the first who gathered some imperfect flowering specimens, which passed into the hands of Professor Grisebach, who referred them, in my ' Bonplandia,' to a new genus of Araliacea (Sciadodendron). The late Mr. Sutton Hayes obtained better specimens of the flowers, which he sent to Eng- land, and which I found without a name both at the British Museum and at Kew. There being no leaves of the .plant in these herbaria, I did not fail, when, in May, 1866, at Panama, to collect specimens of them, so that I am now able to publish a plate of this common, but so little known tree, that Bentham and Hooker, in their ' Genera Plantarum,' were obliged to refer it amongst the doubtful genera. Sciadodendron belongs to the little group of Araliacea proper (dis- tinguished from the Sederacex by the imbricate aestivation of the petals), and it difl'ers from Aralia, Slilbocarpa, and allied genera by its 10-12-merous flowers. XLVI. Sciadodendron, Griseb. in Seem. Bonplandia, 1857, p. 7. Char. gen. emend. ; Calyx limbo undulato v. subintegro. Petala 10-12, ovato-linearia, jestivatione imbricata. Stamina 10-12, antheris incum- ON THE GENUS SCIAUODENDRON. 97 bentibus. Ovarium inferum, subglobosum, 10- v. per excessum 12- loculare, stylis totidem inferne cohaerentibus, superne divergentibus et radiatim expansis abbreviatis. Ovula in loculo solitaria, pendula. Drupa . . .—Arbor excelsa, inermis, glabra, foliis longe petiolatis supradecompositis, foliolis breviter petiolulatis ovatis acuminatis mu- cronato-serrulatis ; iimbellis 15-40-floris compositis, pedicellis ^edun- culisque bractea seariosa subrotunda basi suffultis. Species unica :— 1. S. excehum, Griseb. 1. c. p. 7. (Tab. Nostr. n. 7.) Nomen ver- naculum Panamense " Jobo de lagarto," Nioaraguense "Palo de la- garto."— Geogr. distribution, Isthmus of Panama (Sutton Hayes! Duchassaing ! Seemann !), Nicaragua, between Leon and Ocotal (See- mann!), and Peru (Gay! n. 486). Sciadodendron excehum, the type of the genus, does not seem to be truly indigenous to the Isthmus of Panama ; at all events in that country I have never seen it anywhere but near human habitations. The same remark applies to Carthagena. But I have found it truly wild in the forests of New Segovia, Nicaragua, where it is one of the most common trees, and attains a height of sixty feet. In the Paris herbarium I have seen specimens of it from Peru, where they had been collected by Gay (n. 486). This tree is 60 feet high, with a corky, greyish bark. The branches are few, straight, and terminated by the foliage. The leaves are with- out stipules, from 3-5 feet long, compoundly pinnate, and impart to the tree a Palm-like habit, not unlike that of Caryota. The petiole is jointed at the base of the ramifications of the lea,f, and the joints swollen. On the whole, the leaves resemble those of Eeteropanax fragraus. Seem. The tree flowers without leaves and very seldom, I having seen it but once in flower. The flowers grow from the old wood, and are greenish. In the Isthmus of Panama the tree is known by the name of Jobo de lagarto (AUigator Jobo), from a certain resemblance of the bark of the tree to the skin of an alligator. In Nicaragua it is vernacularly termed " Palo de lagarto " (Alligator's tree), the name of " Jobo " be- ing unknown there. In the isthmus the natives apply the leaves (macerated) with beneficial eft'eot to ulcers. The ashes of the wood are used in Nicaragua in the manufacture of soap. Explanation op Plate VII., representing Sciadodendron excelsum, Griseb. Tlie flowers from specimens collected by Suttoii Hajes, at Panama, and 98 REVISION OF THE NATUEAL ORDER HEDERACEJ!. kindly lent by Dr. Hooker; the leaves from specimens collected by me in the same place. — Fig. 1. Flower. 2. Petals. 3. Flower with petals removed. 4. The same, with stamens removed. 5 and 6. Stamens. 7. Pistil. 8. Cross- section of ovary ; — all magnified. XXIX. On THE Genus Panax. Panax has been made one of the great lumber-rooms of our science, and none of the modern botanists have assigned to it' intelligible limits. Linnaeus refeiTcd three species to it, including two generic types, and his generic name ought to go with the majority of species represented by P. quinquefolium and P. trifolium. Both these species have a dimerous ovary and fire petals, quincuncial in aestivation, the latter character assigning them a place amongst the Araliaceee proper ; whilst Linnseus's third species (P. fruticosuni) has petals valvate in aestivation, and therefore stands as the type of the genus Nothopanax in my Kevision of HederacecB. If the species referred to Panax by authors are examined, it will be found that but few agree generically with P. quinquefolium and P. trifolium ; but that all those that do agi'ee have the same habit and partake of similar medicinal properties, being perennial herbs, with tuberous roots, and with verticillate, digitate leaves, terminal solitary simple umbels, and coloured berries ; more- over, they are distributed geographically over North America and Eastern and Central Asia. I therefore restrict the genus Panax to these species only. It will be seen that these views are at variance with those advanced by Decaisne and Planchon (Revue Hortic. 1854, p. 105), who shift P. quinquefolium and P. trifolium to Aralia, retain the name Panax for P. fruticosum, and unite with it Folyscias, Forst. ; Gheiro- dendron, Nutt. ; Pseudopanax, C. Koch; and Maralia, Pet. Thouars. Bentham and Hooker f. (Genera Plant, p. 938) entertain similar views with regard to shifting P. quinquefolium and P. trifolium to Aralia, the retention of the name Panax for P. fruticosum and its congeners, and the union of Cheirodendron and Maralia with their Panax; but they exclude, as I had previously done, Polyscias and Pseudopanax. Yet, even with these two latter genera removed, Bentham and Hooker's Panax remains ill-defined. Having excluded Pseudopanax, the learned authors ought to have removed Cheiroden- dron along with it, as that genus is so closely allied to it as almost to justify those desirous of forming large genera to unite the two. Oli- ^oscias {Maralia), which the two authors also incorporated with ON THE GENUS VANAX. 99 Panax, has nothing to do with that genus as defined by them, having ruminate albumen, which would entitle it a place among their tribe Hedereai. XLVII. Panax, Linn. Gen. n. 1166, ex parte. — Char. gen. emend. ; Pedicelli articulati. Flores calyculati, polygamo-monoici. Calycis margo obscure 5-dentatus. Petala 5, ovata, acuta, aestivatione quincunciali. Stamina B, filamenta brevissima; autherse ovatse v. oblongfe. Ovarium 2- V. per excessum 3-locuIare, loculis 1-ovulatis. Drupa baccata, didymo- compressa (colorata), 2- v. per excessum 3-loculare, loculis l-spermis. Semina pendula. Albumen sequabile. — Herbse perennes, Amer. bor., Japon., Chin, et Ind. or. indigense, radice crassa, tuberosa ; canle soli- tario, foliolorum verticillum unicum proferente ; foliis palmatim 3-5- foliolatis ; umbellis terminalibus simplicibus. — Aureliana, Lafit. Mem. du Ginseng. Araliae sp. auct. Eupanwe (subgenus), Torrey et Gray, PI. N. Am. i. p. 647. 1. P. trifolium, Linn. Spec. ii. p. 1058 ; Torr. et Gray, Fl. N. Amer. i. p. 648. — P.pudlla, Sims, Bot. Mag. t. 1334. Aralia tri- folia, A. Gray, Bot. N. U. States, ed. v. p. 199. Nom. verriac. Bor.- Amer. "Dwarf Ginseng," v. "Ground Nut." — North America.from Ca- nada to the mountains of the Southern United States (Kalm 1 Bartram ! ann. 1764, Nuttall ! Kertland !). — Leaves sometimes 5-foliolate. 3. P. qmnquefolium, Linn. Spec. ii. p. 1058 ; Torr. et Gray, PL N. Amer. i. p. 648; C. A. Meyer in BuU. de la Class, phys. math. Ac. St. Petersb. i. 840; Walp. Eep. v. p. 925. — Aureliana Canadensis, Lafit. M6m. du Ginseng, cum icone ; Oatesb. Hist. Carol, iii. p. 16. t. 16. Araliastrum foliis ternis quinquepartitis, Ginseng et Nizier officin., Trew, Plant. Select, i. t. 6 ; Lam. Encycl. Me'th. ii. 723. t. 860, fig. 1; Bot. Mag. t. 1333. Aralia guinquefolia, A. Gray, Bot. N. U. States, edit. v. p. 199. Nom. vern. Bor.-Amer. " Ginseng."— Woods of North America, from Canada to the Southern United States (Kalm! Bartram! Kertland!). 8. P. Ginseng, C. A. Mey. 1. c. ; Walp. Eep. v. p. 924.— P. quin- quefolium var. Ginseng, Eegel, ' Gartenflora,' 1862, sp. 314. t. 375 ? P. quinquefolium, p. P. Coreense, Siebold, Verb. Bat. Genoot. xii. P. Schinseng, var. Coreense, Nees, Suppl. Plant. Med. i. t. 16. fig. A.— Mantchuria and Corea. 4. P. Pseudo-Ginseng, Wall. Plant. Asiat. Ear. ii. p. 30. t. 137; C. A. Meyer, 1. c. ; Walp. Eep. v. p. 924 ; Pharra. Centralblatt, 1832, 100 REVISION OF THE NATURAL ORDER HEDERACEjE. p. 353. t. 3. F. Schimeng, var. Nepalensis, Nees, 1. c. fig. C.^Moun- tains of India (Wallich ! n. 3730, Hook. f. et Thomson ! Griffith !) 5. P. Japonicum, C. A. Mey. 1. c. ; Walp. Eep. v. p. 934. — P. quin- quefolium, var. Japonicum, Sieb. P. Schm-seng, var. Japonicum, N. ab Esenb. 1. c. fig. )8. — Japan. 6. P. bipinnatifidum. Seem. (sp. nov.) ; perenne, herbaceum ; foliis verticillatira 3-5-nis, digitatim 5-foliolatis, foliolia bipinnatifidis, supra ad nervos pilis hispidis sparsis, demum glabratis ; pedicellis apice articulatis, calyculo obscuro ; stylis 3-3. — Sanding, East Indies (Herb. Hook.). — "Berries like crabs'-eyes, exactly, with a broad black pallet at top." Allied to P. Pseudo- Ginseng. Species exchisce : — P. aculeatum, Kii.= AcantJwpanax axiuleatum. Seem. P. Anisum, De CaMdi.=Notkopanax Jnisum, Miq. P. anomalmn, H.ook.=Noihopanax anomaliim, Seem. P. arboreum, Yox&i. =^NothopancMO arboreum, Seem. P. armatum, 'Wa\\.=Aralia armata. Seem. P. Australasia, '?exa.=Nothopanax arboreum. Seem. P. ausiralis, Sc]ixiti,.= Sermas australis, Spr. P. attenuatum, '&yiaxi%^=Bidymopanax glabratmn. Dene, et Planch. P. Ujugum, 'Wii\.[.=Pentapanax LeschenauUii, Seem. P. Boivini, 'DcTie.=Sciadopanax Boivini, Seem. P. calvum, C\ia,xa.=DidymMpanax calvum. Dene, et Planch. P. Garibbaum, ^\Q\y.=I)idymopanax glabratum, Dene, et Planch. P. cepMlobotrys, F. M.\xd\.=NotJiopanax cephalobotrys, Seem. P. chrysopliyllum, 'Vah.\=Didymopanax chrysophyllum, Dene, et PL P. chrysopliyllum, Yw^'p.=Bidymopanax Pappigii, Dene, et Planch. P. cochleatum, De CaxiA.^^Nothopanax cochleatum, Miq. P. Colensoi, Ylook. t^=Notkopanax Colensoi, Seem. P. concUfolium, '&oT:!o.=Nothopanax cochleatum, Miq. P. coriaceum, 'Regd—Pseudopanax crassifolium, 0. Koch. P. crassifolium, Dene, et 'Plar^ch.—Pseudopanax crassifolium, C. Koch. P. crenatum,, Dietr.= P. curcifolium, Grifiith= BraMaJo^««s Hainla, Seem. P . decaphyllum, ^&goi=: Sciadophyllum decapJiyllum, Seem. P. decompositum, 'Wal\.=Aralia Oachemirica, Dene. P. dendroid-es, ¥. MueW. =Nothopamx sambucifolium, C. Koch. ON THE GENUS PANAX. 101 P. digitatum, Sweet= P. divaricatwm, Sieb. et Z\icc.= Acanthopanax divaricatum. Seem. P. Edgerleyi, Hook. t—Raukaua Edgerleyi, Seem. P. elegans, 'Fxas.=Not}iopanax elegans. Seem. P. Pinlaysonianum, Wall. =^mZia Mnlaysoniana, Seem. P.foliolosum, Wall.^Aralia foUolosa, Seem. P. Forsteri, Done, et Vla,Ti6h..=Polyscitts pinnaia, Porst. P./rugrans, B.oxh. =Seteropanaixfragratis, Seem. P.fruticosum, Lmn.=Noihopan,ax fruticosum, Miq. P. Gaudichaudii, DeCani.^Cheirodendron, Gaudiohaztdii, Seem. P. glabratum, H. B. Y^.=Bidymopanax glabratum. Dene, et Planch. P. Chinnii, Hook. f.=N'otkopanax Ounnii, Seem. P. ('.i^ Eainla, De Cand.=5ras«ajqpsM Hainla, Seem. P. Heyneanum, 'Wa\\.=Polyscias pinnata, Povst. P. Heyneanum, Wall. Q,&i.=ArthropTiyllum. P. horridum, Smith.=Horsfieldia horrida, Seem. P. innovans, Sieb. et Zucc.=/^caM^Aqpfl«fla; «»ko»(J!MS, Seem. . P. JacManwm, WaR.^ArthrophyUum. P. LescJienauUii, De Ceiadi.=Fentapana!i; Leschenaultii, Seem. P. Lessonii; De Cand.=Pseudopanax Lessonii, C. Koch. P. lineare, Hook. f.=Pseudopanax lineare. Seem. P. longipetiolatum, Vo\i\.=Bidymopanax. P. longissimum. Hook. i.=Pseudopanax crassifolium, C. Koch. P. Loureirianum, De Cand.=.LC