FLORA or THE ISTHMUS OF PANAMA. 143 
In tlie single specimeji I have before me, tlie herbaceous terete and striate stem is prostrate aud 
rootiug, fixed bj thiu rootlets at every interuodc; the rootlets descend iminediatclj below the leaves, 
opposite to their axils, where they arise as little warts, or lenticella). In this manner the plant is enabled 
to fix itself to the aerial roots of the Mangrove-tree {Uhizopliora) . The obovate leaves, 6-8 lines in 
length and almost of the same breadth, are round at the summit, with a somewhat unequal crcnafc margin, 
and attenuated into slender petioles of about the same length; in texture they are rather fle^shy ; tlicir 
midrib and lateral veins are very indistinct. The uppermost leaves are linear, sessile, acute, very short 
and bract-like, sustaining the trichotomoua inflorescence. The involucre of the flower-heads is cylindrical 
or bell-shaped, and the scales of it are in aboiit four or five series, of which those of the bvo outermost series 
are ovate and very short, whilst the scales of the innermost and middle series have a length of 2^-3 lines, 
being oblong or lanceolate. All scales are of a somewhat firm texture, three- to five-nerved, obtuse at the 
apex and much apprcssed. Each involucre contains about ten tubular perfect flowers. The style is most 
singular; it is not thickened at the base, aa in BnclcelUuy Shirrlioj)horus^ and many other CompositijC^ but 
really immersed in a somewhat spongious or fleshy emarginated tuber, which ia separable from tlie style, 
and often broken in four or five pieces when incautiously separated, resembling in some degree the four 
seed-like nucules- surrounding the base of the style in tlie Order Lahiatrn or BQrragine{B. The a<;ha?nia are 
of a yellowisli or fawn colour, somewhat curved aud mostly tetragouous, with obtuse angles, but sometimes 
trigonous with, a convex back. 
Plate XXIX. Fig. 1, au entire capitulimi; 2, receptacle, scales, and a flower; 3, a single flower; 4^ 
the summit of the achseniura ; 5, flower cut open ; and 6, a stamen : — all magnified. 
551. C(ELESTiNA scabrinsciilaj Bcnth, in Orstcd Compositau Ccntro-Amcricana)^ p. 8. no. 27 
(1853)*. Volcano of Cliiiiquij Ycraguas. 
I hare not seen an original specimen, collected by Orsted, but Mr. Bentham's brief and characteristic 
diagnosis enth'oly accorda witb Dr, Secmaun's pluut^ so that there caimot remain any doubt that it belongs 
to the same species- 
552. Adenostemma sp. On road-sides, Panama. 
r 
The genus Adenosfemma needs an entire revision ; many characters given in the diagnoses of the species 
are found to be verj variuhlc, viz. the numher of the cla^ate bristles of tlic pappus and the surface of the 
acha3nia, which I found in the same specimen, the younger ones almost smooth, aud the ripe ones uiuricate. 
Therefore the single specimen collceted b}^ Dr. Seemann, being in a bad condition, is not detenniuable. It 
seems however to be allied to Adenosfemma Idrtijlorum, Bentli. Plant. Uartweg. p. 75. no. 531, of which I 
have compared an original specimen ; but it differs by the tip and the teeth of the leaves being more obtuse, by 
the strongly rauricate ripe achieiiia, and by the clavatc bristles of the pnppns varying in number from three, 
four, and five. On the other hand it has more the habit of ^. Ilumloldlii, C. H. Sehultz Bip., MSS., of which I 
examined a specimen, gathered by Moritz at Jlerida in Teuezucla. But in this species the stem, peduncles, 
and involucre are covered with a dense gland\dar pubescence, whilst in Dr. Seemann's plant the stem and in- 
volucre are ahnost glabrous, the pedicels being covered with short and simple hairs without any gland. In 
the achasnia of both I fouud no striking difference. The plant of Moritz is somewhat stouter. 
553. Hebeclinium macrophi/Uum, Do Gaud. Prodr. torn. 5. p. 136. no. I. — Ageratum caivukum, 
Sicb. Martiiiic. no. 192 ! (non Dcsfont.). Panama, in waste places. 
* Under the name, ' Composita3 centroamcricanae. Bestemmelser eg Beskrivelser af G. Bcntham. Ted 
A. S. Orsted,' has appeared a pamphlet, 49 pages in 8vo, which, being without any otlier title, will, I sup- 
pose, be published in the Transactions of one of the Natural History Societies at Copenhagen, probably in 
the ' A^'idcnskabelige Meddelelser fra den naturhistoriske Forening i Kjobenhavn.' 
