1-28 



Chauiaccristapavonis, long as it looks, is but by one syllable 

 longer than a somewhat recent generic name Pseudocyuwptenis, 

 and is of just the same length as Xcoivasliingtoina, still more 

 recently proposed. 

 Washington, D. C. 



TWO MISINTERPRETED SPECIES OF XYRIS 



By Roland M. IIarpkk 



The name Xyi'is ficxuosa Muhl. has been almost always ap- 

 plied to a certain widely distributed species which is about the 

 only representative of its genus over most of the glaciated region 

 of the northeastern United States.* This name is usually con- 

 sidered as dating from the first edition of Muhlenberg's Catalogue, 

 published in i8 1 3, but in that work there is nothing by which the 

 species can be definitely identified, and indeed no specific descrip- 

 tions were attempted in the whole catalogue. (The words in the 

 fourth column, on which so much stress was laid by Mr. Bick- 

 nell and Dr. Robinson in discussing the identity of certain species 

 oi Agrimonia a few years ago, are expressly stated b\' Muhlen- 

 berg in his preface to be merely the English names of the species, 

 and they cannot therefore be regarded as descriptions.) For the 

 original description of Xyris flex2iosa we must turn to the first 

 part of the first volume of Elliott's Botany of South Carolina and 

 Georgia, published in 18 16, in which four species o{ Xyris were 

 recognized. Two of these were new, based on the collections of 

 Dr. Baldwin in Georgia, and another was identified by Elliott 

 with X. brcvifolia Mx., but was later found by Dr. Chapman to 

 be quite different, and named by him Xyris Elliottii. The remain- 

 ing one is X. Jlcxiiosa Muhl., and the description, habitat, and 

 time of fiowering assigned to it ponit clearly enough to a plant 

 with corkscrew-like stem aiid twisted leaves which we now know 

 to range from New Jersey to Florida and Texas, mostly in the 

 pine-barrens, and which was known to nearly all 19th century 

 authors as X. toria. P^lliott gives as a synonym X. caroliniana 

 Walt., but this s[)ccics can hardly be identified, since it was the 



*.Sce I\l)i)(li)ra 7 : 73. 1905. 



