114 ORCHIBACEM 



often confusing and likely to lead to erroneous conclusions, so 

 that more segregates have been made than is necessary. To de- 

 scribe the species in such a way that determination may be made 

 without hesitation is extremely difficult, as the range of variation, 

 as well as the extent of distribution, is often great ; and species 

 which are quite distinct when widely separated stations alone are 

 considered, run together by imperceptible degrees when specimens 

 from the entire range are compared. While Spiranthes j^rceeox, 

 Watson (of which a full account will be given farther on), tends 

 toward the Mexican S, graminea, Lindl., on the one hand, S. ver- 

 nalis, Eng. & Gray, approaches the Mexican species on the 

 other. 



The S. graminea or S. vernalis group, well characterized by 

 oblong-lanceolate or hnear-lanceolate leaves and by a single-ranked 

 raceme during anthesis, is at present vaguely understood, and con- 

 sequently the species are often confused. From S. vernalis sev- 

 eral segregates have been made which can hardly stand the test 

 of rigid examination, and the following notes, based on my inves- 

 tigations among these segregates, are offered with the hope that 

 they may clear away difficulties. 



Sixty years ago Engelmann and Gray described Spiranthes 

 vernalis in the Boston Journal of Natural History 5 : 236 

 (1845), erecting the species on material collected by F. Lindheimer 

 near Galveston, Texas, on wet prairies in April and May, 1843. 

 The type specimens are fragmentary, though still in a satisfactory 

 condition for determination. They agree with certain other Texan 

 Spiranthes examined, and are not unlike the New England plant 

 described by me as 8. neglecta, though more slender, with nar- 

 rower, longer leaves, with auricled, hyaHne-margined floral bracts, 

 and with a slightly different hp. One of the specimens is very 

 slender, and though devoid of flowers is without doubt pure S. 

 vernalis, agreeing with its companions so far as vegetative parts 

 are concerned. The collection on which S. vernalis was based 

 comprised at least two species, one clearly referable to Spiran- 

 thes prcecoXf Watson. I have examined four sheets of this col- 



