208 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



be the species best known. In the " N. American Flora," Rose's 

 descriptions of praealtum and confusum are evidently copied from 

 .Hemsley, "Biol. Centr. Americana." and he even says of the former, 

 " a very doubtful species." His description of dendroideum, on the 

 other hand, is clearly original, though whether based on living or 

 dried material does not appear. 



The descriptions are not sufficiently full to point to satisfactory 

 distinguishing characters of flower between this species and the other 

 two, but it appears easily separated by its leaves (see fig. 119), which 



Fig. 118. — S. dendroideum M09. and Sesse. 



are. distinctly stalked and have a nearly orbicular lamina. In my 

 plant, also, they have a whitish margin when young (due to a waxy 

 secretion) and a purplish margin when old, while in the two other 

 species the leaves are wholly green ; the leaf also is of much firmer 

 texture than in praealtum or confusum, and the stem is stiff and 

 erect, with few ascending branches, while in the others it branches 

 frequently and soon forms a low bush. The petals, described as 

 lanceolate, appear to be broader than those of praealtum, narrower 

 than those of confusum. The inflorescence is large like that of 

 praealtum, not congested as in confusum. In the following description 

 the stem and leaf characters are taken mainly from my living plant, 

 the remainder from De Candolle, Hemsley, and Rose. To judge 



