WESTERN TROPICAL AMERICA, 171 
The above description is taken as well from Mr. Hinds’s specimen, as from 
another gathered by Hartweg near Bogota. It is very distinct from all the 
species I am acquainted with, but agrees in several points with the character 
given by Miquel of his H. Richardiana. The leaves and spathe are, however, 
longer, the bracts or spathilla much broader, and the stalk scarcely thickened 
under the spathz. There is also a H. flexuosa of Presl, but I have not access to 
any description of it. 
828. Heticonta vaginalis, sp. n., foliis caulinis oblongo-lanceolatis utrinque longe angustatis, 
petiolo fere ad apicem vaginante, spathis (4) lanceolatis acuminatis complicatis 8-12-floris, spathillis 
lanceolatis, perigonii laciniis interioribus usque ad apicem concretis.—Isle of Gorgona. 
Folia tria adsunt, inferius sesquipedale, superius 7-pollicare, omnia 14-2 poll. lata, vaginis elongatis 
scapum arcte involventibus, lamina fere ad vaginam decurrente, apice longe (nec ut in 1. acuminata abrupte) 
acuminata. Scapus e folio superiore breviter exsertus. Rhachis 3-pollicaris, flexuosus. Spathe in specimine 
4, infima 5 poll., suprema 2 poll. longa, omnes extus leviter puberule. Spathille acutissimee, membranacee, 
complicatee, exteriores sesquipollicares.  Pedicelli fructiferi e spatha exserti. Perigonium sesquipollicare, 
undique coloratum; foliolum exterius infimum canaliculatum, 2 superiora interioribus alte adnata et paullo 
breviora ; interiora connata in unum acutissimum integerrimum. Genitalia non vidi. Capsula baccata, fere 
hemispheerica, truncata, 5 lin. lata. 
829. Canna, sp. A. glauce affinis—Guayaquil. 
This is the same as Hartweg’s n. 698, from the same locality. The species 
of this genus have been so multiplied and distinguished by characters, of so little 
value or fixity, that it is impossible to determine these from dried specimens, 
until the whole genus shall have been worked up by a competent botanist. 
830. Maranta arundinacea, Linn.—Panama. 
It is surprising that, after the structure of the flower in this and allied 
genera has been so well explained by Nees and others, the outer sterile peta- 
loid stamens should be still described in general works as inner lobes of the 
corolla. The ovary is, in a young stalk, three-celled, although, even at the time 
of flowering, two of the cells are already much smaller and empty. 
831. RENEALMIA racemosa, Poepp. et Endl. Nov. Gen. et Sp. 2. p. 26.—Caulis 8-pedalis. 
Racemus compositus, fere glaber, ramis 3-5-floris.—Salango. 
In one of the flowers which I examined of this plant, the central lobe of the 
so-called labellum of some botanists, or synéme of Lestiboudois, was converted into 
two perfect stamens, the anthers being borne on short filaments, with a dilated 
connectivum, one perfect cell, similar to the larger one of the stamen usually 
fertile, and one smaller rudimentary cell. The two lateral lobes of the syneme 
were larger than usual, oblong, but contracted at the base; there was no trace of 
the sixth stamen, whose place would be between the pair of antheriferous lobes. 
