530 PROCEEDINGS OF THE ACADEMY OF [PaRT III 



1. Castilleja sessiliflora Pursh. 



Castilleja sessiliflora Pursh, Fl. Am. Sept. 738. 1814. "In upper Louisiana. 



Bradbury v.s. in Herb. Bradbury." Imperfectly described. Better 



characterized by Nuttall, who knew the plant growing, as Euchroma grandi- 

 flora, Gen. N. Am. PI. 2: 55. 1818. 



Corolla yellowish-green, the posterior lip pink-margined, the 

 anterior yellowish-white and only slightly greenish-thickened at 

 base. Bracts and calyx distinctly yellowish. Flowering in May 

 and June. 



Sandy or rocky plains and bluffs, escarpments of the Staked 

 Plains of northwestern Texas; Upper Sonoran Zone (at altitudes 

 above 1500 meters). Through the Great Plains from Illinois and 

 Manitoba to Wyoming and northwestern Texas. 



Texas. Hall: EstelUne, Reverchon 4316 (M, U, Y). Randall: 

 Canyon, Pennell 10542 (A, H, M, 0,.T, U, Y). 



2. Castilleja labiata Pennell, sp. nov. 



Root perennial. Stems several, 1,5— 2 dm. tall, arachnoid-lanate 

 with close pubescence. Leaves 3-7 cm. long, linear to lanceolate, 

 slightly 3-ribbed, trifid, with narrow lobes, finely cinereous -pubescent 

 especially on the lower surface, green above, paler beneath. Bracts 

 broader, conspicuously trifid, probably distally yellowish or grean- 

 ish-yellow. Sepals 22-25 mm. long, those of each side united f 

 length, acutish to acuminate, puberulent, apparently yellowish. Co- 

 rolla 35-40 mm. long; posterior lip 10-12 mm. long, with thin mar- 

 gins; anterior 5-7 mm. long, proximally deep-green, distally flaring, 

 evidently yellow or yellowish; anterior lobes glandular-pubescent, 

 posterior with longer glandless hairs. Stigma about 1 mm. wide. 

 Capsule not seen. 



Type, stony prairie (uncommon), San Angelo, Texas, collected 

 in flower May 19, 1899, W. L. Bray 353, in United States National 

 Herbarium; isotype in Herb. University of Texas. 



I regret that the spring of 1920 had been so dry, and consequently 

 so early, that on my visit to San Angelo on May 18 and 19, all 

 Castillejas were in fruit and shriveled, I failed to find this species, 

 but the plant I did find and which appeared to be common proved 

 to be C. citrina, below. Though dishking to base a species of this 

 critical genus on so few collections, I think C. labiata must be 

 recognized. It appears to be a primitive member of the alliance of 

 C. purpurea (Nutt.) G. Don, connecting that group with C. sessili- 

 flora Pursh. 



Texas. Tom Green: San Angelo, Bray 353 (T, U). Valverde: 

 Devil's River, Pilsbry (A). 



