ORCHIDACE^ 



grains are simple, so that this character would seem to be con- 

 stant in the Eupogonia group and may well be regarded as a 

 weighty bit of evidence in favor of restricting to Pogonia only 

 those species of the alliance which produce simple rather than 

 compound pollen grains. It is true that Wettstein in a paper on 

 Cephalanthera, Epipactis and Limodorum/ in which he advo- 

 cated the union of these genera under Epipactis, dismissed the evi- 

 dence of simple pollen grains in Cephalanthera on the ground that 

 in this genus they simply break down earlier than in Epipactis. 

 Col. M. J. Godfery^ in a discussion of this paper maintains that 

 Wettstein has treated the subject of the pollen grains in a some- 

 what disingenuous manner and that, notwithstanding his conclu- 

 sions, the differences between the pollen grains of Epipactis and 

 Cephalanthera pointed out by Richard are undiminished in irn- 

 portance. In other words, Godfery is inclined to view the nature 

 of the pollen as important in systematic work among the orchids, 

 at least with regard to Epipactis and Cephalanthera. 



Pogonia, Cleistes, Psilochilus, Isotria and Triphora are mem- 

 bers of that large section of the Orchidaceae which is character- 

 ized by mealy or powdery pollen, a section recognized by L. C. 

 Richard and recently adopted by Dr. Rudolf Schlechter as the 

 Polychondreae. Together with Pogoniopsis, Monophyllorchis, 

 Xerorchis, Vanilla, Galeola, Eriaxis, Epistephium and Lecanor- 

 chis they constitute the Vanilleae. 



If we examine Triphora, Pogonia, Cleistes, Isotria and Psilo- 

 chilus with a view to finding key characters to separate them 

 clearly from one another, the following would stand out as the 

 most weighty ; 



1 Oesterreiche hotanische Zeitschrift 39 (1889) 395-399, 4^22-430. 

 '^Journal of Botany 5S (1920) 69-74. 



[6] 



