454 Britton: Studies of West Indian plants 



tube; stamens shorter than the corolla; fruit obpyramidal, 3-4 mm, 

 long, 1.5 mm. thiek at the top, tapering gradually to the base, 

 densely pubescent, the persistent calyx-lobes about 0.5 mm. long. 



Limestone rocks, northern parts of Havanna and Pinar de. 

 Rio. Type from coastal hillside, Bay of Mariel, Pinar del Rio 

 (Britton & Earle 7619). 



A barren specimen from limestone rocks at Cape Corrientes 

 (Britton & Cowell 9894) is doubtfully referred to this species. 



48. THE GENUS HEPTANTHUS Griseb. 



A Cuban genus, of which three species were described at the 

 place of original publication (Cat. PI. Cub. 148. 1866); all were 

 from western Cuba, H. cochlearifolius Griseb. being typical, and 

 none have since been added to the genus. I now propose three 

 others, all from eastern Cuba. 



Leaf-blades glabrous above or merely puberulent. 



Peduncles longer than the leaves. 1. H. cochlearifolius. 



Peduncles shorter than the leaves. 



Leaf-blades orbicular-ovate, 1-2 cm. long; petioles 



densely villous. 2. H. cordifolius. 



Leaf-blades triangular- ovate, 4-6 mm. long; petioles 



sparingly villous. 3- H. Shaferi. 



Leaf-blades pubescent on both surfaces. 



Leaf-blades repand-dentate; peduncles much shorter than 



the petioles. 4- H. brevipes. 



Leaf blades incised-serrate or lobed; peduncles as long as 

 the leaves or longer. 

 Leaf-blades 12 mm. long or less, incised-serrate. 5. H. ranunculoides. 



Leaf-blades 1-2.5 c m - long, deeply 3-lobed, the lobes 



coarsely few-toothed. 6. H. lobaius. 



I. Heptanthus cochlearifolius Griseb. Cat. PI. Cub. 148. 



1866 



Type locality: Western Cuba. 



Distribution: Pine-lands and savannas, Pinar del Rio. 



The upper leaf-surfaces are glabrous or nearly so. 



A plant related in leaf-form, but with the upper leaf-surfaces 

 densely puberulent is common in pine-lands and white sand in the 

 central districts of the Isle of Pines, but could not be found in 

 flower during our visit to this region in February and March, 1916 

 (Britton & Wilson 14163) ; it is therefore referred to this species, 

 with doubt. 



