nerved lip. The Costa Rican C. gracilipes Schltr. has 
broader leaves, similarly smaller sepals and differently 
shaped, 5-nerved petals. 
Ayacucno: Aina, between Huanta and Rio Apurimac, at 750-1000 
meters altitude, epiphyte in open woods, May 7, 17, 1929, E. P. 
Killip & A. C. Smith 23154 (Tyee in Herb, Ames No. 61960). 
Cryptocentrum peruvianum (Cogn.) C. Schwein- 
Surth comb, nov. 
Centroglossa ? peruviana Cogniaux in Bull. Soc. Roy. 
Bot. Belg. 48 (1906) 331. 
An examination of the type material of Cogniaux’ 
concept, borrowed from the Herbier du Jardin Botani- 
que de V’Etat at Brussels, shows conclusively that it is 
very dissimilar to the genus Centroglossa as illustrated 
by Cogniaux (in Martius FI. Bras. 3, pt. 6 (1904) t. 46, 
figs. 1 and 2). It should be referred to the genus Cryp- 
tocentrum, bearing the combination above. 
The combination Cryptocentrum peruvianum was used 
by Schlechter (in Fedde Repert. Beihefte 27 (1924) 108) 
in the discussion of C. Hoppii. No description was given 
and no other occurrence of the combination has been 
found in literature. Consequently the Schlechter usage 
is anomen and cannot affect the combination here made. 
Contrary to the description, the sepals and petals are 
acute or acuminate, and neither obtuse nor truncate—a 
characterization doubtless due to the fact that in the type 
specimen these organs usually appear to be broken off 
at the tip. In addition, the lip, like the entire perianth, 
is fleshy and not membranaceous as described, and it is 
lanceolate in outline, rather than ovate-rhomboid. 
Cryptocentrum peruvianum has two rather close allies. 
It appears to differ from the Colombian Cryptocentrum 
Hoppii Schltr. (which I have not seen) in having broader 
leaves, purple (not greenish yellow) flowers and a longer 
[ 188 | 
