morphic species of this alliance, is scarcely of generic 
weight when taken alone. 
Finally, the factor of an ascending rostellum appears 
to be a matter of degree, even if not too recondite and 
obscure for ordinary recognition in the dried specimen. 
It appears to me that the wise course is to relegate 
Petalocentrum to the synonymy of the variable genus 
Sigmatostala. 
The concept, Petalocentrum angustifolum, the only 
one of the genus not previously described as representing 
Sigmatostalix, was considered identical with Sigmatosta- 
liv pusilla Schltr. (cf. Kriinzlin in Engler Pflanzenr. LV. 
50, pt. 2 (Heft 80) (1922) 312), and the generic identity 
of the concepts Petalocentrum and Sigmatostalia was 
once suspected by Kriinzlin (ef. lc. 318). 
Sigmatostalix peruviana /folfe in Kew Bull. 
(1910) 371. 
Sigmatostalix pusilla Schlechter in Fedde Repert. 10 
(1912) 392. 
Sigmatostalix bicornuta Rolfe in Kew Bull. (1913) 842. 
Petalocentrum pusillum Schlechter in Fedde Repert. 
15 (1918) 145; Krinzlin in Engler Pflanzenr. IV. 50, 
pt. 2 (Heft 80) (1922) 312; ex Mansfeld in Fedde 
Repert. Beih. 58 (1980) t. 59, nr. 285. 
Petalocentrum angustifolum Schlechter in Fedde Re- 
pert. 15 (1918) 145; ex Mansfeld in Fedde Repert. 
Beih. 58 (1930) t. 59, nr. 234. 
Petalocentrum bicornutum Schlechter in Fedde Re- 
pert. Beih. 9 (1921) 179. 
All of the concepts cited above appear to be closely 
similar vegetatively, although Sigmatostalia pusilla is 
described as having markedly shorter pseudobulbs and 
leaves than the others. It is significant that the sepals 
and petals of all of the concepts are of almost exactly the 
[ 76 | 
