900 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
RANGE: In damp or wet woods, Panama to Venezuela. 
ILLUSTRATION: Engl. & Prantl, Pflanzenfam, 4°: fig. 46, L, M. 
SPECIMENS EXAMINED: ; 
PANAMA: Loma de la Gloria, near Fat6 (Nombre de Dios), Province of 
Colon, near sea level, Pitticr 4081. Forests around Puerto Obaldia, 
San Blas Coast, near sea level, Pittier 4293. 
This species is distinguished from all others so far found in Panama by the 
presence of two small but conspicuous stipitate glands at the base of the ovary. 
It is upon this character that Karsten based his generic name Biglandularia. 
It is very probable that when more complete material of the various members 
of the genus has been secured, it will be found that Disadena, which antedates 
Biglandularia, is a valid genus. Four species of this section are known, but 
they do not form a homogeneous group. One of them has no calyx, while of 
the other three two have appendaged anthers and one has unappendaged ones. 
Leiphaimos azureus is a slender plant with few or numerous slender, very 
succulent stems, each of which is furnished with two or several pairs of thin 
bracts. The roots often form a dense mass, and are fleshy and knotted. The 
flowers are small, the corolla limb being about 1 em. broad, resembling those 
of some species of Primula. The resemblance of the corolla to those of some 
of the primroses led Baker to apply the name primuloides to one of the African 
species,” 
A NOTE CONCERNING THE GENUS RANDIA, WITH DESCRIPTIONS 
OF NEW SPECIES. 
The genus Randia is one of the larger groups of the family Rubi- 
aceae and is represented in North America by about 40 species. Other 
members of the genus occur in ‘South America, and a still larger 
number in the tropics of the Old World. The group is not a par- 
ticularly well-marked one, being very closely related to Gardenia. 
Indeed, most of the genera of the tribe Gardenieae are seperated by 
rather artificial characters, 
In 1873 Hooker? removed from Randia a group of American 
species, associating them in a new genus which he named Basana- 
cantha. Randia was limited to the species with perfect flowers, 
the flowers of Basanacantha being dioecious. The latter group was 
characterized also by certain habital peculiarities, none of them of 
very great importance. As the genus Basanacantha was originally 
delimited it included a homogeneous group of species, but Urban 
later added two West Indian plants of very different habital char- 
acters, plants which in general appearance are very like the common 
species of Randia proper. After study of all the North American 
species it seems to the writer that the two genera are separated by 
too artificial a character, and that they should be united. The spe- 
cies of Basanacantha listed below are, therefore, transferred to the 
genus Randia. For the other North American species of the group 
the proper combinations have already been made. 
* Kew Bull. 1894: 26, 1894, *In Benth. & Hook. Gen. Pl, 2: 82. 
