February 29, 1912 5 



E. aliena: i specimen, the type. 



E. grisea: 2 specimens, Leiberg ^^j6 and Barber ^j. 



E. cariescens: 2 specimens, Brewer 2064, Heller yi^Q, both 

 unjointed, but Heller's plant with a few of the style tips broken 

 off. 



This ends the record as to the "unjointed and persistent" 

 character of the style in the material before me, and it may 

 easily be seen that there is but little in the way to permit an 

 arrangement of the named forms into two groups, which, for 

 present convenience, may be called the Ciliatae and Triflorae 

 respectively. 



Color oj the petals. — Now let us give attention to this phase 

 of the subject, first reviewing what literature leads us to expect. 

 After the Latin description of G. ciliatuju, Pursh (1) says: "flow- 

 ers resembling those of No. 6." No. 6 is G. rivale^ and of that 

 he says "flowers yellowish-purple." According to various sub- 

 sequent authors the "flowers" or "petals" are: "white," Nuttall 

 (2); "yellowish-white, tinged with purple, persistent," Torrey(4); 

 "purplish," Gray (5), Brewer and Watson (8); "richer purple 

 than common," Cooper (6); "grande.s, blanche veinees de pour- 

 pre," Provancher (7). Of these I especially note with pleasure 

 the statements of Torrey and Provancher; but let me also call 

 attention to the works of more recent authors, where we will 

 find the color recorded as "light purple," Britton (10), Rydberg 

 (14); "pale purple," Piper (13); "purplish," Robinson and Fer- 

 nald (15), Coulter and Nelson (16). 



Greene (J 2), however, says of his E. campamilata^ "petals 

 crimson tinted; of E. dissecta^ "petals yellow," and of E. Jiavula^ 

 "calyx-segments purple-veined like the petals and yellowish," 

 but of the remainder of his species the question of the color of 

 the corolla is left for one to guess at or find out as best one may. 



As stated in the introduction to the paper, the petals in the 

 northern Utah plants collected by me v^'qxq pale yellow^ more or 

 less margined and veined with dull reddish^ a combination that 

 I failed to interpret as "purplish." This coloration being well 

 preserved in my specimens, I have used the same as the basis 

 for comparison in making the tabulations, from which the fol- 

 lowing records are drawn: 



