The American Species of Limnorchis and Piperia, north of Mexico 



By Per Axel Rydberg 



In revising the manuscript of the Orchidaceae for Dr. Brit- 

 ton's Manual of the Flora of the Northern States and at the same 

 time that of my Catalogue of the Flora of Montana and the Yel- 

 lowstone Park,* I came to the conclusion that the genus Habe- 

 naria as treated in America was an altogether too diversified 

 genus and contained species of very distant relationship. In 

 Europe, as a rule, several genera are admitted. Prof. E. Pfitzer, 

 who prepared the Orchidaceae for Engler and Prantl's Die natiir- 

 Hchen Pflanzenfamilien, places the North American Habenariae in 

 the following genera : Corfoglossum, Gymnadcnia, PlatantJiera, 

 Perulana and Habenaria. Kraenzlin in his OrcJiidaeearnni Cen- 

 tra et Species, admits Habenaria, Gymnadcnia and PlatantJiera. 



Habenaria in a separate subtribe from 



the rest. The true Haben 



M 



Central America. They are characterized by the long filiform 

 appendages of the stigmas, anthers and petals. PlatantJiera as 

 understood by Pfitzer is still a very complex genus and had he 

 known the American forms as well as those represented in Europe, 

 he undoubtedly would have restored Rafinesque's BlcpJiar glottis for 

 our fringed orchids and established a few more genera. During 

 last July I stayed a few days in London and met Mr. Rolfe, the 

 authority on orchids at Kew. He expressed the opinion that 

 Habenaria must be split up, not only so that Gvmnadenia and 

 PlatantJiera should be taken out, but that all three genera should 

 °e subdivided, In one instance he went farther than I had dared, 

 v 'iz., expressing the opinion that Orchis rotundifolia perhaps should 

 be removed from Orchis, 



When doing my work referred to above, I came to the conclu- 

 sl on, that if Coeloglossnm and Peru/aria should be admitted, certain 



represented wholly by American species must be 



other groups 



Mem. N. Y. Bot Garden, Vol. I. 



605 



