CONTRIBUTIONS TO WESTERN BOTANY. 637 



of it. It is the Astragalus trijiorus of Watson in King's 

 Rep., at least in part, of Coulter's Manual, and Of most 

 western collectors, and it may prove to be only an extreme 

 form of y4. Wethei'illi Jones, in which case the latter 

 name will' prevail, but at present it seems sufficiently dis- 

 tinct. It seems to be confined to the mountainous regions 

 of Colorado and adjacent States and Territories. It 

 blooms throughout the season, often in bloom quite late. 



Astragalus trijiorus Gray, 1. c. This seems to be close 

 to A. Candolleanus (H. B. K.) Sheldon, and being the 

 first name in the genus, should stand. It is not at all 

 certain that it is distinct frorn Phaca trijiora DC. Ast. 

 62, t. I, 1802. It seems to be a very variable species. 

 The typical form is annual, with the habit of A. Geyeri 

 Gray, and seems to be an early bloomer. Specimens be- 

 longing here are C. Wright, New Mexico, the smallest 

 specimen on the sheet from the Mexican Boundary Sur- 

 vey; one sheet of Dr. Mearns, Carrigallilo Mountains, 

 New Mexico, April 18, 1892. A taller form (which may 

 be A. cerussatus Sheldon) from El Paso, Texas, G. R. 

 Vasey, is apparently perennial, and has purple flowers. 



The various forms of this species which may deserve 

 varietal rank are A. trifeorus var. Candolleanus (H. 

 B. K. Nov. Gen. et Sp. 6; 495. 1823), with oblique pods 

 and many leaflets (usually 8 to 15 pairs). 



A. TRiFLORUs var. iNSULARis (Kellogg, Bull. Cal. Acad. 

 Sci. i; 6. 1884), with few leaflets and scarcely oblique 

 small pods. It is quite possible that this will still hold as 

 a good spe'cies. 



Astragalus Pondii Greene is close to the two above 

 varieties and may deserve specific rank. It was poorly 

 described, but authentic specimens show differences. It 

 is the second specimen on the sheet in Cal. Acad. Sci. 

 along with A. insular is, described by me in Contributions 

 4, p. 28, last form described. 



