282 IOWA ACADKMY OF SCIENCES. 



Distribution. — Emmet County. (Herb. S. U. I.) Little 

 Rock, Spirit Lake, C. R. Ball. 



CNICUS UNDULATUS, A. Gray. 



Cnicus undulatus, A. Gray. Proc. Am. Acad. 10: 42. 

 1874. 



. A. Gray. Syn. Fl. N. Am. 1 : 403. 1884. 



Cirsium undulatum, Spreng. Syst. Veg. 3: 374. 



. DeCandolle, Prodr. 6: 651. 1887. 



. Torrey & Gray, Fl. N. Am. 2: 456. 1843. 



. Gray Man. 273. 1868. 5 Ed. 



Carduus undulatus, Nutt. Gen. 2: 130. 1818. 



. Greene. Proc. Acad. Nat. Sci. 1892: 359. 



. Britton & Brown. Illustr. Fl. N. St. 3: 486. 



/ 4063. 1898. 



Low biennial two to four feet high, white woolly, branches bear- 

 ing a single head with purple flowers; stem striate, white woolly, 

 or somewhat glabrate below; leaves, the lower eight inches to a 

 foot or more long, undulate, lobed or pinnatitid, with prickly den- 

 tate lobes, upper surface at first arachnoid later smoothish; lower 

 densely canescent with prominent veins connecting with spinescent 

 tips, upper sessile, lanceolate or oblong-lanceolate, deeply pinnati- 

 fid, to spinescently lobed; heads one and three-fourths to two inches 

 high, flowers purple; bracts of the involucre appressed, arachnoid, 

 outer ovate to lanceolate with a rather weak prickle and a thick glu- 

 tinous ridge on back, inner bracts long, lanceolate acuminate, straw 

 colored and serrate, flowers purple, corolla tube ten lines long, the 

 corolla lobes four lines long, tips clarate, anther tips strongly 

 acute, filaments pubescent; achenium smooth, pappus of outer 

 flowers merely barbellate, the inner plumose. 



This is an extremely variable species. Its distribution is 

 said to be from Lake Huron, calcareous islands to the 

 Northwest territory and to British Columbia and Oregon. 

 The species is certainly not common east of the Missouri. 

 In company with Mr. Cratty the writer found what 

 undoubtedly should be referred to his species in Kossuth 

 County. No. 607 (I. S. C). 



This is not typical for the species, but in its habit of 

 growth and character of leaves more nearly approaches 

 this than C. canescens. It is certainly not C. altissimus 

 wOsY filipendnlus. 



