lo Muhlenberg-ia, Volume 8 



To quote further from Greene (13 174), in Sieverda the 

 "mature styles are long, filiform, plumose to the very apex." 

 And again (13 175), referring to "the type that has been called 

 Geiim tj'ifioriun and Sieversia triflora,^'' "the accrescent and 

 plumose character of the styles it has in common with Siever- 

 sia.'''' Now the fact is that the style is accrescent only to the 

 elbow^ hook, or articulation (or where such would be if present), 

 and only the accrescent part is red a.nd plumose, the non-accres- 

 cent terminal portion being T^xs^ciicoWy glabrous Sind light brown, 

 i)ersistent or deciduous. Only when the terminal portion is 

 gone from the style, can that organ be said to be "plumose 

 throughout." But such a statement must indeed be seriously 

 questioned as having taxonomic value here. Others of the 

 characters of Erythroconia, however, are the erectness of the 

 petals and calyx lobes, color of calyx and corolla, and, I may 

 add, flowers nodding in authesis; but all these characters are 

 common to Geum Hvale, which really seems to be closer, to my 

 mind, to Erythrocoma than does Acomastylis. It seems to me, 

 therefore, that Erythrocoma must be given due place as a sec- 

 tion of Geum, sejisu latu, and distinct from Sieversia; or G. 

 rivale must be made the type of another genus, co-ordinate with 

 Sieversia, Acomastylis, Erythrocoma, etc., in which case the 

 real status of Caryophyllata Tourn. must be investigated. 



Conclusions. — While certain questions that have arisen dur- 

 ing this study must be left for further investigation, I feel that 

 I am justified in coming to the following rather definite conclu- 

 sions. Geum ciliatum Pursh, and G. trifiorum Pursh, are 

 worthy of recognition as separate species, neither name being 

 synonymous with the other. Probably these two species are the 

 only forms worthy of specific standing, the other named forms 

 being better considered as varieties of one or ihe other. Ornata, 

 dissecta, campanulata, and perhaps brevifolia, seem to be recog- 

 nizable varieties of E. ciliata. I have failed to find characters 

 sufficiently strong to really separate y?flZ7^/« and australis from 

 the variations existing; in the aeorreo-ale which I, as well as Dr. 

 Greene (12 177), "fail to resolve into its specific elements by 

 anv certain characters." 



