deep blood-red. Ligules pink, white for half the length. Witoto 
name: mu-se-na. Spanish name: cacao silvestre.’? June 20, 1942, 
Richard Evans Schultes 4010 (Tyrer in Herb. Gray). 
Herrania nitida (Poepp.) R. E. Schultes var. as- 
pera (Karsten & Triana) R. EF. Schultes comb. nov. 
Brotobroma aspera Karsten & Triana ex Triana Nuev. 
jén. e esp. plant. fl. Neo-Granad. (1854) 12. 
Herrania aspera (Karsten & Triana) Karsten in Lin- 
naea 28 (1857) 447, 
Arbuscula parva ab Herrania nitida principaliter foli- 
olis utrinque asperioribus et apicem versus margine defi- 
nite sinuatis, florum partibus omnibus sanguineo-pur- 
pureis, fructibus probabiliter minoribus differt. 
With the searcity of collections, it is rather difficult 
to evaluate this concept. It is certainly not specifically 
distinct from the widespread and somewhat variable 
Amazonian Herrania nitida. At the present state of our 
knowledge, it is probably best treated as a variety. The 
very definitely and regularly sinuate margin of the upper 
part of the leaflets is apparently a constant character, 
although in several of the Peruvian collections which I 
have assigned to Herrania nitida there is a tendency for 
the margin to be obscurely undulate. 
Karsten, in discussing Herrania aspera (in Linnaea 
loc. cit.) and in pointing out its difference from HZ. pul- 
cherrima Goudot, stated that it occurs ‘‘in vallis Oreno- 
censis marginibus ad pedem Andium bogotensium mer- 
idensiumque. ... et littora fluminis Magdalenae.’’ He 
cited no specimens. The presumed type of the concept 
— Triana 5333—was collected in the Llanos of Villavi- 
cencio. The only other collections which have as yet 
come to my attention are a topotype, one from the 
Apaporis basin and several from the trapécio amazénico. 
Therefore, | believe that Karsten’s assertion that Her- 
rania nitida var. aspera is found in the Magdalena basin 
[ 130 | 
