February 29, 191 2 9 



men, however, would warrant placiuj^ it with E. nffinis. Strange 

 that with so much material, none is at hand from east of Illi- 

 nois. Newfoundland and Labrador specimens ouj^dit to be 

 worthy of special study. 



The sheet of McCalla 2oyj in the Cornell Herbarium in- 

 cludes two quite different thin<^s, the one on the left being a 

 true co-type of E. ajjinis^ the other a fair example of E. ciliata. 

 Persons having specimens of this collection should note care- 

 fully the figures in plate j. Also, collectors who may visit 

 Banff, Alberta, may find field observations there to be esiK-cially 

 interesting. 



The specimen of Ilartitian joo from Chihuahua, Mexico, 

 seems to represent the Mexican form of E. trifiora^ but it will 

 probably prove to be indistinguishable from E. arizonica. It 

 bears the same relation to E. arizonica that E. aliena does to E. 

 grisea (see key). Barber 57 from Williams, Arizona, is evi- 

 dently E. grisea^ tho apparently not determined by Dr. Greene. 



Final discussion. — Dr. Greene, under E. ciliata (12 177) 

 says: "Nothing is more easy than to distinguish these plants as a 

 groujD, from the eastern E. trijlora. To this I can readily agree, 

 tho perhaps with a trifle less emphasis. To me this distinction 

 can readily be extended to include all of the named forms, giv- 

 ing us the groups Ciliatae and Trijlorac, as designated above. 

 Now if these groups are indeed capable of separation on these or 

 other characters, what about the long-accepted merging of the 

 two original species of Pursh? And again, if these two groups, 

 or species, or what one may choose to call them, be realh' sep- 

 arable on the characters presented in this paper, what about the 

 continued inclusion of the Ciliatae in the genus Sicversia of 

 Willdenow, or in the section Sieversia under Geitin, as accepted 

 by Gray? Surely a readjustment, or recharacterizalion of some- 

 thing should be done. 



As claimed by Greene, these plants seem to be representa- 

 tive of a generic type sufficiently distinct from both topical 

 Geiim and Sieversia, and his <^tx\\\9, Erythrocoma must therefore 

 be accepted, if such a view is to be acknowledged. Erythrocoma^ 

 however, must be, in part at least, recharacterized, and careful 

 analysis of the matter meets with important difficulties. 



