always have dry, dehiscent fruits with prominently winged, scle- 

 rotic seeds. Epistephium smilacifolium Rchb.f. from New Cale- 

 donia has justly been elevated to a genus of its own by Halle 

 (1977) because of the scandent habit of the plants and their 

 unique, three-locular ovary with axile placentation in addition 

 to its geographical separation. I have personally studied plants 

 of this species in the field in New Caledonia together with those 

 of Eriaxis. The plants of this latter genus, however, contrary to 

 the statements of Burns-Balogh and Funk, have a unilocular 

 capsule with parietal placentation. The genus Galeola itself in its 

 present status is an assemblage of strikingly discordant elements. 

 Having removed Cyrtosia, I also find it necessary to reinstate 

 from synonymy the genus Erythrorchis on account of the plants 

 being saprophytic, the long and slender column of the flowers 

 provided with a short but distinct foot, and the nature of the 



column 



m 



with large foliaceous bracts; the long, slender footless column of 

 the flowers is basally fused with the lip to form a small, saccate 

 to subtubular nectary. For these plants I propose a new genus 

 based on their twining habit and Vanilla-like flowers in addition 

 to the above mentioned characters. 



Systematic considerations. 



In reassessing the genera here grouped around Vanilla as was 



com 



ponent parts in the past were quite distantly removed from one 

 another. Both Rhizanthella and Palmorchis are currently 

 regarded as representatives of their own subtribes. Their actual 



m 



subfamilies 



thus, expressing not even a most tenuous relationship between 



them 



matic 



cations were considered. A review of the various individual 

 approaches not attempted here, because in principle such an 



228 



