Inflati 105 



long, the general outline nearly oval, narrowly sulcate ventrally, most- 

 ly translucent. Flowers normally white, inclined to be cleistogamous, 

 about 5-7 mm. long, the petals mostly but little longer than calyx. 

 Banner notched, oval 2-4 mm. long, with sides little arched along 

 the edges, about as long as keel or a trifle more. Wings oblong, 

 nearly as long as banner. Keel very wide relatively, about 1.5 mm. 

 wide below and at ti]) abruptly rounded to erect and but a little higher 

 than the rest of keel and square. Calyx campanulate, the tube about 

 2 mm. long, the teeth arched, triangular and fully as long as tube. 

 Pedicels stout, 1-2 mm. long, shorter than the delicate bracts. Ped- 

 uncles 1-3 cm. long, the floral rachis about as long. Leaves 5-7 cm. 

 long, ascending, on short petioles. Leaflets 5-8 pairs, folded, not con- 

 tiguous, oblong, rounded at tip, about 1 cm. long. Stipules acuminate 

 from a deltoid base. Stems 1-2 ft. long, rather many from the crown 

 and branched below, with internodes shorter than the leaves. Pubes- 

 cence soft throughout, spreading, hoary. The flowers vary greatly in 

 color and length of petals. The banner is often arched to erect, with 

 light-pink claw and blade striate-purple-veined and with darker edges, 

 the groove is a half circle and occupies most of banner or broadly 

 triangular, in some specimens the banner has a purple streak below 

 the notch and on the s'des. The wina;s equMl the kepi and n'^rending 

 with tips just touching the keel tip, obliquely ovate and flat and obtuse. 

 The keel tip is acute and narrow, pink but not purple-tipned. The 

 wings are often light-pink and often innple strenked, usually darker 

 below. Other material has dark-purple flowers with very striate ban- 

 ner. Some Colorado forms have the banner 2-3 mm. longer than k^el 

 and wings about intermediate. Frequent in the Lower Temperate life 

 zone from Baker Lemhi Co. Tdnho and the San Rafael Swell TTtah to 

 centra] Colorado and southward at least to central Mexico. The type 

 locality is Mexico not Peru. Some f'^rms are short-lived rerennia'ts. 

 Blooms in summer Tt grows in the hills and not on the plains in this 

 country. DeCandolle's fanciful figure is poor, that of H. B. K. is better. 

 On the plains it is replaced by the following. Gray's type is mac' e up 

 of A. lentiginosus var. diphysus partly. 



The type of Phaca triflora DC. is a fanciful fisure drawn as though 

 it were a weak annual, but corresponds in essential featurps with the 

 species. Phaca Candolleana H. B. K. which w^s intended to he thp 

 same thing is drawn without the root as though it were a perennial and 

 has narrower leaflets but otherwise the same. Tt also is a fanciful 

 figire. Plants from the Valley of Mexico by Schaffner are clearlv an- 

 ual and early blooming and fit DeCandolle's figure. Plants of my 

 own collection at Casialidad Zacatecas and Ramos correspond well 

 ■^■^■ith both fig\ires showins the species to be a winter annual or flower- 

 ing in the spring and fall from the same plant, a thing very common 

 In Mexico, and plants insei)arable from the Colorado form called cenis- 

 satus by Sheldon. The type is too n^ar to cerussatus for it to be 

 kept up as a variety. Plants corresponding to cerussatus from Mexico 

 also are those of Rose from Irola No. 4558. At Baker Idaho I found it 

 growing and in fruit in September and evidently in condition to live 

 over the coming winter and thus at least a biennial. Parish sends me 

 a form from Leastalk San Bernardino Co. California clearly peren- 

 nial with mottled smooth pods and short calyx teeth. Plants growing 

 in the San Rafael Swell Utah have the dense hnb't of cerussatus and 

 fine and soft pubescence and beautifully mottled nearly globose pods, 

 elongated calyx teeth and dark-purnle flowers, appearing so like A. 

 pubentissimus that only an examination of the pods and calyx teeth 

 can sei)arate them. These plants also bloom as annuals and winter 

 -annuals but the two species d-^ nnf seem to hybridize. A. Coquim- 

 bensis answers fairly well to this species. The species is very 

 variable in the pods and pubescence. 



