CHATANTHERA PERUVIANA (sp. nov.): annua, tenella, diffuse ramosa; foliis lineari-cuneatis vel spathulatis versus apicem spinulosodentatis laxe villosis mox glabratis, summis angustioribus circa capitulum confertis; involucri squamis subscariosis retusis, exterioribus ovalibus, costa in appendicem nunc folioformem producta, interioribus lineari-oblongis sæpe mucronulatis; ligulis linearibus fere glabris involucrum vix superantibus, labio interiori parvo brevi apice bidentato. Andes of Peru, above Baños. Near C. tenella: the first species detected north of Chili.

ORIASTRUM COCHLEARIFOLIUM (sp. nov.): pulvinatum, laxe arachnoideo-lanatum; foliis in caules breves confertis imbricatis sessilibus crassis obtusissimis muticis dorso mox glabratis intus sub margine incurvo concavis lanuginosis, inferioribus oblongis, superioribus spathulatis capitulum sessile arcte rosulato-cingentibus; involucri squamis omnibus scariosis, apice radiante colorato ovato-lanceolato acuto rigidiori; pappi setis capillaribus rigidis basim versus parce barbellulatis superne fere lævibus. Alpamarca, high Andes of Peru. A very distinct species, interesting from its extending the range of the genus farther north than before. Ray-flowers perhaps fertile; their linear ligule obscurely tridenticulate at the apex, and with two minute teeth at the base on the inner side, representing the lower lip. Mature achenia unknown. Those of O. Chilense, Wedd., are pyriform; the papillæ of their surface when soaked swell into a jelly, and then the achenium appears as if glabrous. Its pappus in the ray-flowers, generally of two or three caducous bristles, is sometimes wholly wanting; that of the disk-flowers is nearly uniserial, the bristles united at the base into a ring. They are finer and softer than in O. pusillum; but it is not worth while on this account to keep up Aldunatea as a section. — 0. pusillum has abortive stamens in the ray-flowers, not before noticed, still more approximating the genus to Tylloma and Egania, which last might well enough be referred to Oriastrum, and even both, perhaps, back to Tylloma.

A serious error in transcription vitiates Weddell's amended character of Oriastrum, i. e. the achenia of the disk, instead of those of the ray, are said to be glabrous and effete; those of the ray, instead of those of the disk, papillose and fertile.

JUNGIA FERRUGINEA (Linn. f.): scandens vel sarmentosa; foliis 5-9-lobatis subtus pannoso-villosis; capitulis 5-10-floris glomerulatis, glomerulis in corymbos vel thyrsos congestis; squamis involucri

interioribus paleisque arcte involutis flores et pappum subæquantibus; acheniis glabris. Bogota, Mutis? Holton. Quito, Jameson, Couthouy, &c.

JUNGIA PANICULATA (Dumerilia paniculata, DC. Jungia ferruginea, Don et auct., non Linn. f. J. spectabilis, Less., non Don.): fruticosa; foliis subtus tomentosis, tomento albido implexo; capitulis conferte cymosis plerisque pedicellatis multifloris; involucri squamis interioribus paleisque floribus "luteis" pappoque subdimidio brevioribus ; acheniis pilosiusculis. Petioli nunc nudi nunc basi quasi stipulati. Peru. I suppose (although I cannot now verify the supposition) that Linnæus received his J. ferruginea, along with most of the new species from "America Meridionali" described in the Supplement, from Mutis, therefore probably from Santa Fé de Bogota, where Dr. Holton collected what is manifestly the Linnæan species. In this species the individual heads, only 5-10-flowered, are commonly so closely clustered in fascicles as to explain, if not to justify, the view taken by the younger Linnæus of a compound capitulum. The Peruvian species referred by Don to J. ferruginea is quite different. De Candolle's (but not Don's) J. spectabilis is the same as his Dumerilia paniculata without the stipular appendages, which are inconstant.

PEREZIA. Dr. Schultz goes too far when he refers the Mexican and North American species of this extensive genus to Trixis. In the former even the fewest-flowered species have a gradated imbricate involucre and erostrate achenia. The latter has a uniserial involucre, the scales all of the same length, with or without a circle of spreading, mostly foliaceous bracts.

To Trixis frutescens I refer T. paradoxa, Cass., T. cacalioides, Don, and T. Neæana, DC. T. angustifolia, DC., which is probably a narrow-leaved form of the older T. corymbosa, Don, is known by the linearlanceolate scales of the involucre gradually tapering to a point, the margins of the leaves usually revolute. T. obvallata, Hook. & Arn. probably belongs to T. longifolia, Don.

Cichoracea.

ACHYROPHORUS CHONDRILLOIDES (Oreophila chondrilloides, Don in herb. Hook. Seriola Brasiliensis, subvar. b., Hook. & Arn. Comp. Bot. Mag. I. p. 30): glaucescens, undique glaberrimus, radice fusiformi; caule folioso stricto mono-oligocephalo; pedunculis elongatis; foliis subcarnosis lineari-lanceolatis integerrimis seu obsoletissime

denticulatis, superioribus subamplexicaulibus, imis in petiolum basi dilatatum sensim angustatis; involucri squamis lanceolatis subacutis. -Rio Negro, North Patagonia, in saline soil.

ACHYROPHORUS SESSILIFLORUS (A. Quitensis, Schultz Bip. Wedd. with A. Humboldtii and albiflorus, Schultz), a widely variable species, must include not only A. sonchoides, DC. (the most caulescent form), but also a

Var. B. BARBATUS (A. barbatus, Schultz Bip. Rev. Crit.): minor; involucri phyllis exterioribus superne pl. m. setosis.

Var. γ. SUBRUNCINATA (A. setosus, Wedd. & A. eriolanus, Schultz Bip.): foliis runcinato-dentatis vel incisis margine sæpius setuloso-ciliatis; involucri phyllis exterioribus oblongis seu obovatis dorso setosis vel nudis. Ludit, 1, involucro tomentoso, 2, foliis rhombeo-ovatis longius petiolatis.

ACHYROPHORUS STENOCEPHALUS, Gray (including A. taraxacoides Wedd.) is perhaps only an extreme variety of the preceding. Meyen's specific name was taraxacifolia, which Walpers, perhaps accidentally, changed to taraxacoides, which name both Weddell and Schultz cite under A. Meyenianus (which is most probably a form of A. sessiliflorus) as well as under the present species, showing some confusion, to avoid which I have retained the appropriate name of A. stenocephalus. PICROSIA LONGIFOLIA, Don. The pappus is fulvous and soft, not fragile, and the genus is probably nearest related to Pyrrhopappus.

FITCHIA NUTANS, Hook. f. Of this curious arborescent Cichoracea Professor Dana collected a single specimen on the mountains of Tahiti, which is about 25 degrees of longitude farther east than Elizabeth Island, where it was discovered by Mr. Cuming. The single capitulum in the collection being male adds nothing for the completion of the character of the genus. The plant from which it was taken is said to be a tree, with yellow flowers.

2. Notes on Lobeliacea, Goodeniacea, &c. of the Collection of the U. S. South Pacific Exploring Expedition. By Asa GRAY.

Lobeliacea Sandwicenses. The Sandwich Islands are remarkable for their arborescent, shrubby, or fleshy-stemmed Lobeliacea. The species are numerous and peculiar, but very difficult to investigate in herbaria, owing to the imperfection of materials in collections and to

the injuries from insects to which these and other lactescent plants are especially liable. There are moderately good materials extant in different collections of ten or eleven species, and indications of almost an equal number; while many others doubtless remain to reward the labors of future explorers of the forest region of Hawaii, a large part of which lies still untrodden by the naturalist. Exclusive of three true Lobelias, and of a striking new Isotoma? of Kauai or Nihau in Remy's collection, the known species of the Sandwich Islands may all be referred to Gaudichaud's genera Delissea, Cyanea, and Clermontia, three genera which also shade off into each other in a somewhat troublesome manner. The only essential character of Gaudichaud's genus Rollandia, viz. the adnation of the stamineal tube with one side of the tube of the corolla, is as I suppose a mistake. At least it does not occur organically in flowers of the plants which well accord with the (now flowerless) specimen of R. lanceolata collected in Freycinet's voyage, upon which Gaudichaud founded the genus, nor, I believe, in the plant which answers to his more miserable specimen of R. crispa. The former is a good Delissea; the latter, having larger and somewhat foliaceous calyx-lobes, is one of the species through which Delissea graduates into Cyanea. To the latter genus we may confidently refer Presl's Macrochilus (Lobelia?) superba, Cham., of which the calyxlobes are probably incorrectly said to be imbricated in æstivation, and also a new and most remarkable arborescent species, which by its extremely long and apparently petaloid calyx-lobes, equalling the corolla in length, approaches Clermontia; but these divisions are perfectly separate down to the ovary, almost filiform, spreading in anthesis, and not deciduous. Our Delisseæ are:

1. DELISSEA LANCEOLATA. Rollandia lanceolata, Gaud. Bot. Voy. Freyc. R. montana on the plate, the upper leaves reduced in size. R. lanceolata var. grandifolia, A. DC. Prodr. is really just the type of the species, which Gaudichaud characterizes as having "foliis magnis."

2. DELISSEA CLERMONTIOIDES, Gaud. Bot. Voy. Bonite, t. 47, which may probably also be D. Kunthiana, t. 77, and even Rollandia Humboldtiana, of the same author, t. 76.

3. DELISSEA DELESSERTIANA. Rollandia Delessertiana, Gaud. 1. c. t. 75. We have what may be a variety of this species, pinnatiloba, from Kauai.

4. DELISSEA CORIACEA (sp. nov.): fruticosa, glabra; foliis amplis (pedalibus et ultra) oblongo-lanceolatis coriaceis repando-serrulatis basi

acutis longiuscule petiolatis, venulis conspicue reticulatis; racemis plurifloris petiolum haud superantibus; calycis limbo obsoleto seu dentibus 5 minutis instructo; corolla pollicari subcurvata. Kauai, Remy.

Var. B. foliis spathulato-lanceolatis in petiolum brevem longe attenuatis. Crater of East Maui. Fruit as large as a cherry.

5. DELISSEA OBTUSA (sp. nov.): suffruticosa; ramis junioribus floribusque undique pubescentibus; foliis (5-6-pollic.) membranaceis oblongis serrulatis apice vel utrinque obtusis subtus parce pubescentibus; racemis plurifloris petiolum gracilem haud superantibus; calycis limbo fere obsoleto; corolla gracili subpollicari incurva. - Mountains of Maui. Var.? MOLLIS caule crassiori; foliis elongatis (subpedalibus) oblongo-lanceolatis basi in petiolum breviusculum attenuatis supra puberulis subtus molliter pubescentibus; "floribus pollicaribus crassiusculis cæruleis.". - Mouna Kea, IIawaii. Possibly both may be varieties of the following.

6. DELISSEA ACUMINATA, Gaud. Bot. Freyc. p. 457, t. 76. Oahu. Var. ANGUSTIFOLIA : foliis elongato-lanceolatis aut angustatis aut latiusculis. D. (Lobelia) angustifolia, Cham., DC. - Oahu.

7. DELISSEA UNDULATA, Gaud., to which belongs D. subcordata of the same work; leaves with the base subcordate, obtuse, or acute being found on the same stem. The small protuberances on the tube of the corolla represented by Gaudichaud occur in all the forms, but are in

constant.

8. DELISSEA? PLATYPHYLLA (sp. nov.): caule fruticoso orgyali petiolisque tuberculis aculeisve conicis mollibus obsitis; foliis sesqui – bipedalibus obovato-oblongis repandis membranaceis glabris; pedunculis axillaribus brevibus crassis paucifloris: lobis calycis glabri brevissimis subulatis. - District of Puna, Hawaii. The port is rather that of Cyanea, and the resemblance to Gaudichaud's Rollandia crispa is not

remote.

To Cyanea, Gaud., distinguished by the foliaceous or enlarged and persistent lobes of the calyx, I refer all the following:

1. CYANEA GRIMESIANA, Gaud. 1. c. t. 75. - Oahu. The corolla is variously stated to be "bluish rose-color," or "white striped with reddish-purple externally."

Var.? CITRULLIFOLIA: foliis bipinnatipartitis, segmentis sinuatis; caule aculeis conicis creberrimis horrido. Mouna Roa and Mouna Kea, Hawaii. Flowers unknown.

2. CYANEA ASPERA (sp. nov.): foliis oblongo-ovatis acuminatis