242 ON SOUTH-AMEEICAN APOCTNACEiE. 



branch bearing many congested flowers; corolla salver-shaped, red, 1 in. long; tube 

 elongated, somewhat twisted ; segments roundly ovate, reflexed ; stamens exserted. 



Thenabbia. 



This genus was established by Kunth in 1818, with a good figure and an ample de- 

 scription. It has numerous large flowers on a slender axillary peduncle, and many 

 long slender pedicels aggregated into an umbellate corymb : the flower has 5 lanceo- 

 late sepals, and a corolla with an extremely short narrow tube and a border of 5 gib- 

 bously ovoid segments, patently expanded, with a simple dextrorse convolution ; the 

 stamens are long and whoUy exserted. The chief peculiarity of the genus consists in a 

 syngenesious union of its broad membranaceous filaments into a tube in the middle, 

 while their extremities are quite free ; they are seated in the mouth of the short tube of 

 the corolla, and are attached above to the anthers ; these have a long indurated con- 

 nective with 2 long basal prongs, and terminated above by a soft triangular expansion, 

 all valvately conniving at the apex to form the stegium which covers the stigma. The 

 style is slender and terminated by the incrassate clavuncle, which is pentagonal and 

 nectariferous, peltate below, with a membranaceous lacinulate margin, and is surmounted 

 by a globose stigma, in accordance with the usual structure of the family; this was 

 not understood by Kunth, who figured the adherent ruptured portions of the base of the 

 clavuncle as expanded extensions of the filaments — a mistake arising from its appearance 

 in the dried flower. Kunth did not see its bifolUcular fruit ; this was described by De 

 CandoUe from a Mexican drawing, where they are fusiform, 6 in. long, 4-5 lines broad, 

 containing many obovoid seeds with an apical coma. 



1. Thenardia floribunda, H. B. K. Nov. Gfen. iii. p. 210, tab. 240; A. DC. Prodr. viii. p. 425. In 



Mexico indigena et in hort. cult. : non vidi. 



2. Thenardia ? stjateolens, Mart. & Galeotti, Bull. Acad. Brux. xi. p. 359 ; Walp. Rep. vi. p. 473. In 



MexicOj prov. Mechoacan : non vidi. 

 A doubtful species. 



The two plants from Guiana, placed in Thenardia by Mr. Bentham, belong to his genus 

 Thyrsanthus. 



PORSTERONIA. 



Under Thyrsanthus it is related how that genus has been confounded with Forste- 

 ronia by Miiller, and their numerous respective species promiscuously intermingled by 

 him. Meyer, in founding the genus Forsteronia in 1818, before the establishment of 

 Thyrsanthus, equally confounded together the few species enumerated by him, so that 

 his generic character is not exact. De CandoUe in 1844 offered a much better diagnosis 

 than that of Miiller (1864); but the following appears to me a still more correct 

 resumen of its characters. There is a remarkable degree of similarity in the external 

 habit of the plants and in the character of the inflorescence in the two genera, the 

 flowers in both cases being too minute to render their difference easily perceptible ; but 

 there can be no mistake in the respective characters of the follicles and seeds. 



