Rydberg: Studies on the Rocky Mountain flora 473 



almost erect branches above; basal leaves glabrous, firm, 1-2 cm. 

 long, pinnately lobed, with lanceolate cuspidate-tipped lobes; 

 stem-leaves sessile and partly clasping, lanceolate, sharply dentate 

 with cuspidate teeth or entire; calyx-tube campanulate, 2 mm. 

 long, somewhat scarious in the sinuses, sparingly glandular- 

 puberulent; teeth subulate, i mm. long; corolla trumpet-shaped, 

 7-8 mm. long; tube nearly twice as long as the calyx; capsule 

 exceeding the calyx; seeds 5 or 6 in each cell. 



This is related to G. sinuata, but differs in the simple, straw- 

 colored, essentially glabrous stem, the glabrous, pale green leaves, 

 and the form and toothing of the stem-leaves. 



Utah: St. George, 1877, E. Palmer 325 * in part (type, in herb. 

 Columbia Univ.) ; also 326. 



Dr. Brand reduced Gilia Tweedyi Rydb. to a variety of G. 

 minutiflora, without having seen a specimen. I do not think that 

 it is rational to do so, for in G. Tweedyi the pod is not i -seeded, but 

 bears 1-3, usually 2, seeds in each cell, i. e., it is 4-seeded. The 

 plant is more closely related to G. inconspicua. 



Dr. Brand made Gilia Haydeni A. Gray a variety of G. subnuda 

 Torr., and gave G. Crandallii Rydb. as a synonym of this variety. 

 Professor Nelson regarded G. Crandallii as the same as G. subnuda 

 and gave G. Haydeni and G. superba Eastw. as varieties of the 

 same. Both arrangements are incorrect. G. superba Eastw. is 

 the typical G. subnuda Torr. characterized by the orange or 

 scarlet corolla. In both G. Haydeni and G. Crandallii the corolla 

 is rose-colored. They are quite distinct from G. subnuda and in 

 my opinion distinct from each other. 



Dr. Brand's treatment of Leptodactylon pungens (Torr.) Nutt. 

 or Gilia pungens (Torr.) Benth. and its relatives is far from satis- 

 factory. He divides it in two subspecies: subsp. A. eu-pungens 

 and subsp. B. Hallii (Gilia Hallii Parish), which is a matter of 

 taste, but he also divides the former in four varieties: a. Hookeri 

 (Dougl.) Brand; b. caespitosa (Nutt.) Brand; c. tenuiloba (Parish) 

 Brand; and d. devestita Brand. 



The first variety is based on Gilia Hookeri Benth. To make 

 this species a variety of G. caespitosa could be also passed over, as 



* The same number in the herbarium of the New York Botanical Garden is 

 entirely different, and belongs to Gilia hutchinsifolia. Maybe some mixing of the 

 labels occurred. 



