WOOTON AND STANDLEY FLORA OF NEW MEXICO. 357 



campanulate or tubular, sometimes gibbous, the teeth mostly subulate or triangular; 

 corolla longer than the calyx, the banner and wings usually exceeding the obtuse 

 keel; banner usually with its side reflexed, occasionally spreading; stamens diadel- 

 phous; fruit a few to many-seeded pod, variously inflated, curved, flattened, or 

 grooved, with one or both sutures more or less inflexed, becoming in some cases com- 

 pletely 2-celled. 



Mature pods are usually necessary for the determination of the species. 



In the consideration of this genus we have been unable to follow those who have 

 separated it into several genera, because the lines of separation do not appear to us to 

 be sufficiently distinct. Acting upon this judgment, we have considered all the 

 species under one generic name. We fully appreciate the merits of Dr. P. A. Ryd- 

 berg's treatment of the species here referred to Astragalus, in the Flora of Colorado,' to 

 which in our own work we are much indebted. WMle we believe thoroughly in the 

 principle of the segregation of species into small genera wherever possible, the small 

 groups must, in our judgment, possess resemblances in more than one set of characters 

 and their members should be readily recognized as close relatives. The characters of 

 the fruit are especially unsuited for the piurpose of generic distinction in this family. 

 As a means of making Doctor Rydberg's valuable work easily usable in connection 

 with our key, we insert his generic names in the key. 



By strict rules of priority, if all the plants here referred to Astragalus are placed in 

 one genus it should bear the name Phaca, that having precedence in Linnaeus's Spe- 

 cies Plantarum over Astragalus. However, it seems to us that there are extreme 

 cases where such a rule may be disregarded and we have preferred to leave the trans- 

 ference of the species to Phaca to some other writer, if it should ever be considered 

 necessary. 



KEY TO THE SPECIES. 



Pods completely 2-celled. 



Pod fleshy. (Geoprumnon.) ^ 1. A. crassicarpus. 



Pods not fleshy. 



Pods inflated-membranaceous. (Cystium) 2. A. diphysus. 



Pods not inflated . 



Pods linear-oblong, somewhat flattened laterally, membranous. (Ha- 



MOSA.) 



Stems very short and stout; plants appressed-sericeous. 



3. A. calycosus. 

 Stems slender, elongated, spreading; plants villous, merely canes- 

 cent, or sometimes glabrate 4. A. nuttallianus. 



Pods ovoid, never flattened laterally, sometimes compressed dorso-ven- 

 trally, coriaceous. (Astragalus.) 

 Mature pods glabrous. 



Flowers purple; stems very short, almost wanting. 



5. A. mollissimus. 

 Flowers yellow or greenish; stems stout, 30 to 60 cm. high. 



Leaves glabrous above, merely strigose beneath; flowers 



greenish yellow 6. A. oreophilus. 



Leaves densely villous, almost tomentose, on both surfaces; 

 flowers bright yellow 7. A. yaquianus. 



^Colo. Agr. Exp. Sta. Bull. 100: 202-212. 1906. 



^ This and the following names inserted in the key are considered as representing 

 separate genera by Doctor Rydberg, in the Flora of Colorado. By us they are regarded 

 as sections. 



