February 29, 191 2 3 



based upon imich of tlic material considered In- Dr. Greene, cx- 

 cepling, particularly, lliat of his private herbarium. 



Hypotheses adopted. — (icitw ciliatum was made known to 

 science by Pursh (1) in 1814. Subsequently, in the supplement 

 at the end of the same work (*4: 736), he described G. tHJlorum. 

 Comparison of the orii^^inal descriptions furnishes little to show 

 that the two are different, or one and the same species; but nev- 

 ertheless the descriptions are suflficient to make me feel, as pre- 

 sumably did Dr. Greene, that real representatives of Pursh's spe- 

 cies are in this collection, the his type specimens have not been 

 seen. 



For purposes of this study, therefore, it is assumed that 

 these two species of Pursh, as weii as all of Greene's segregates, 

 are valid species, and I have accordingly followed the late clas- 

 sification of Greene (1 2) in laying out the material for study 

 and in recording most of my ob.servations. Each sheet of speci- 

 mens w^as handled separately and with care, a good hand lens 

 being used whenever styles were present. Tabulations were 

 accordingly made, especially as to style characters and color of 

 the petals, and such drawings made as seemed advisable. 



Characters of the style. — The genus Sieversia was estab- 

 lished by Willdenow about a hundred years since, and soon 

 Pursh's Geum ciliatum and G. trijioruui were transferred to it 

 by European botanists. Altho Sieversia has not been allowed 

 generic rank by American authors until recently, it was accepted 

 by Gra\-, at least as early as 1856 (5), as a section of Geum, the 

 section being there characterized, in part, as follows: ''Styles not 

 jointed, wholly persistent and straight;" and this is the treat- 

 ment, to-day, in the latest edition of the Gray manuals (15). 

 Also, in the Flore Canadienne (7) I find: ''Styles non articules 

 et persistant en entier. Sieversia R. Br.;" and in Britton and 

 Brown (10): "Style . . strongly plumose throughout in fruit, 

 not jointed." This, however, applied specifically to G. ciliatum. 



But in 1899 Dr. Greene (1 1) reclaimed for Sieversia generic 

 rank, and from him I now quote: "The essential characters of 

 Sieversia as distinguished from Geum are these of its style; this 



