﻿6 n 



making the leaf appear almost laciniate. I have seen no 

 true forms of the typical species in the United States, 

 except a few found along the eastern base of the Rocky 

 Mountains, in Colorado and Wyoming. 

 Arabis arcuata var. perennans (Pringle, Watson, P. 

 A. A. 22, 467). 

 It is impossible to separate this from typical A . arcuata, 

 though extremes seem sufficiently distinct. Coville's No. 

 1747 is surely a reduced form of this; root leaves oblan- 

 ceolate, entire, 6-8" long, rosulate; stem leaves oblong to 

 oblanceolate, sessile, or nearly so, acutish, reduced above; 

 racemes 3-5' long, rather closely flowered; pods a little 

 arcuate, and below horizontal ' on recurved pedicels, 

 1-1 l /2 r long, 1" wide, barely acute; pedicels slender, 2" 

 long; seeds obscurely in two rows, narrowly winged 

 above and not at all at very base, oval, %" long; plants 

 cagspitose and many stemmed. Watson's specimen from 

 the Uintas, at 12,000° alt., is taller and more lax, with 

 longer root leaves, otherwise about the same. Rose's 

 specimen from the Yellowstone Park is the same. My 

 No. 5330, from Marysvale, Utah, May 31, 1895, 6ooo° 

 alt., in gravel, has longer pedicels and wingless seeds, but 

 I fail to find any valid separating character. 



(Ledeb. Fl. Ross. 1, 



There is no character assigned by Watson for this 

 species which holds; generally the pods are linear-oblong 

 and acute at each end, usually curved most on the outer 

 side, and about as long as the pedicel, glabrous, but all 

 these vary, some pods are half the pedicel or even a third 

 the pedicel, others are longer; some are oblong-ovate, 

 and some are blunt at apex, all ours are annuals. Draba 

 hirta, at least specimens so named, from Spitzbergen, is 

 perennial with twisted pods. 



