196 
imens since collected from that region; it would appear, 
Wallichian s 
In alli ye 
? 
is the leaf of quite another species which proves to be identical 
with the entire specimens bearing Wallich’s number 1590, 
preserved at Kew and at Edinburgh. 
The specimens according to the label were collected at Pundua 
by De Silva, who certainly travelled in Silhet and not in 
India, and they agree very well with Roxburgh’s drawing of 
S. colubrina (Wight’s Icones t. 434). 
As a further indication that Silhet is really the habitat of these 
specimens it is of interest to find that on the Edinburgh sheet of 
Wallich, 1590 an inflorescence of S. Wallichiana (an undoubted 
Silhet plant) has been mounted among the leaves. Some 
correlative evidence is also afforded by the sheet of no. 1590 at 
the British Museum, which consists of a fragment of S. 
chi. ch of S. aenea, var. 
S. Bourdilloni, Brandis, which is referred here to S. cinna- 
momifolia var. Wightit, was described from specimens with 
flowers in the bud stage and before the corolla tube had 
developed. 
It seems better to refer the Indian specimens of S. cinnamomt- 
folia to a variety, since they differ from the Ceylon plant in 
having broader leaves with more rounded bases and somewhat 
different seeds. : 
resemblance between this species and S. Nua-vomica 1s 
considerable, but the latter being a tree with different leaves, 
uits and seeds, can easily be distinguished from the huge 
climbing plant. 
68. S$. rupicola, Pierre ex Dop in Bull. Soc. Bot. Fr. lvii. 
Mém. 19 (1910), p. 17, et in Lecomte, Flor. Gen. Ind.-Chin. 1v- 
p. 168. — 
Inpo-Curva. Cambodia. Prov. Samrong tong; Monts Pra, 
Pierre 3688. 
