144 DR. GRISEBACH ON COUTOUBEA VOLUDBILIS, 
lanceolatis acutis tubum ovatum equantibus, corolle *'rubelle" 
tubo exserto tenui lobis obovato-oblongis longiore, capsula ovoideo- 
oblonga, uniloculari: placentis intus non prominulis. 
Herba spithamea, annua; folia inferiora 8" longa, 2"' lata, internodiis 
longiora, superiora distantia, internodiis crescentibus ultrapollicaribus; 
pedunculi sepe tequilongi cymam racemiformem a medio caule con- 
stituentes; calyx 13", corolle tubus 3" longus; anthera ovales, 
incumbentes, paullo exsertz, filamento tenui; stylus ovario brevior, 
stigmate late capitato; capsula 3" longa. 
Hab. In Guatemala, pr. Las Nubes (Wendl., mense Januar.); forma 
minus elongata in vulcano Frasu, Costariez, alt. 9000 ped. ( Wendl., 
m. April.). 
Sect. 3. Microcala, Lk. Calyx 4-dentatus. 
6. C. filiformis, Reichenb. 
7. C. quadrangularis, Gr. 
Of Lisianthus, sect. Brachycodon, Benth., I possess his L. pu- 
milus, which proves a congener of Pagea; and probably L. ramo- 
sissimus, Benth., is P. Poeppigit itself. Mr. Bentham observed the 
anthers to be at length recurved : hence there remain, to distinguish 
it from Lisianthus, the higher insertion of the stamens, the form of 
the corolla, and the very different habit. In the true Lisianthi, 
the enlargement of the connective on the back of the anther-cells 
is peculiar; and this character, if compared throughout the genus, 
may perhaps be of some value in the discrimination of Pagea. In 
the section Helia (viz. in L. brevifolius and L. chelonoides), the 
structure of the anthers is the same as in Chelonanthus and Ma- 
erocarpea: in both species they are at length recurved, and in . 
the former the connective is apiculate: hence the character of 
Helia is chiefly confined to a marcescent corolla, and less peculiar 
than was supposed. A double placenta in each capsule-cell occurs 
likewise in L. chelonoides (a really annual species, from Kegel's 
specimens) and in Z. alatus, Aubl, to which I reduce my L. 
Oerstedii. If this identification proves correct, Aublet’s species 
must be transposed to Helia. L. tetragonus and L. auriculatus, 
Benth., have been reduced by their author to L. acutangulus, Bot. 
Mag., which is L. trifidus, Kth., but not L. fistulosus, Poir., the 
latter, from an authentie specimen, having purple flowers. 
The sections Chelonanthus and Macrocarpea are to be united, 
being only distinguished by the shrubby growth of the latter: the 
shape of the capsule proves of no sectional importance. 
My supposition that Symbolanthus is little distinguished from 
the section Leiothamnus is confirmed by a beautiful Lisianthus 
