SOUTHERN CALIFORNIA ACADEMY OF SCIENCES. 107 



of March of the present year I had a fresh opportunity ot siud\- 

 ing 6". variabilis in Salt River Valley. It certainly deserves its 

 name, for almost every plant seems to have marked peculiarities. 

 Here are some notes I made in the field : 



(i.) Tempe, Ariz. The style-branches and stigmas deep 



crimson; anthers grey, pollen olive-green. The flowers 



seem not to open so wide as 5. fendleri. 



(2.) Phoenix, Ariz. Flowers paler than in the Tempe plant 



just described; styles and stigmas pale purplish-pink; 



anthers and pollen yellow. L,eaves greyish, tripartite, 



the lateral lobes separated down to the base; median lobes 



about 45 mm. long and 22 broad; lateral lobes about 22 



mm. long and 19 broad; all coarsely crenate. Green 



fruit shows no signs of cusps; it is maliform, densely 



stellate-pubescent; sides of carpels reticulate. This may 



stand as form tripJiylla; it is parallel with 6". incana form 



dissecta (var. dissecta, Gray, PI. Wright, i. 21). 



(3.) Phoenix, Ariz. Styles and stigmas whitish, with a faint 



purplish tint: anthers with deep purple lobes; pollen p le 



yellow; petals short, long. 11, lat. 10 mm., vermillion; 



leaves long and narrow, wavy-margined, with basal lobes; 



fruit about as in f. tripliylla. 



(4.) Phoenix, Ariz. Styles and stigmas deep crimson ; 



anthers dark crimson, becoming black at maturity; pollen 



yellow; flowers bright vermillion; petals long. 15, lat. 13 



mm.; leaves fairly broad, tri-lobed. 



Growing mixed with the S. variabilis^ both at Tempe and 



Phoenix, I found Malvastrinn Cctilieri, Watson, with deep 



orange flowers — an exquisitely beautiful plant. This has been 



referred to Sphcvralcea, but it has the fruit of a Malvastriun, and 



belongs there if there is an}' validity in the characters used to 



separate these genera. 



There is no doubt that ^. variabilis is closelj^ related to 

 .S. fendleri, and perhaps it should stand as a subspecies of it. 

 Then Xh^ fendleri series will be classified thus: — 

 (i.) Sphoeralcea fendleri, Gray. This is the form with deeplv 

 lobed rather short leaves; I have a tracing of the type, 

 kindly sent by Dr. B. L,. Robinson. The flowers are 

 bright red in life, not " rose-red". The plant is common 

 in New Mexico, from Las Vegas and Santa Fe to the ^ 

 Mesilla Valley. 

 (2.) Spliceralcea fendleri f. lobata (5". lobata, Wooton, Bull. 

 I'orr. Bot. Club, xxv. 306). This differs in the form of the 

 leaves, which are longer and larger; it is really comm')ner 

 than XxM^ fendleri, with which it completely intergrades. 

 Professor Wooton informed me, however, that the 5. 



