360 PROCEEDINGS OF THE ACADEMY OF [1892. 



islands in Lake Huron," and botanists of the middle sections of the 

 Union would do well to collect again the type. Its real characters 

 cannot be made out from Nuttall's diagnosis, except as compared 

 with eastern species. Gray's description does not pretend to 

 define the plant of any one locality, but is made loose enough to 

 cover that vast aggregate of specimens which he had from all parts 

 of the country between the Great Lakes and the Pacific, and from 

 Manitoba to Mexico. Such a collection exhibits forms very distinct 

 in general appearance, and enough of them to make five or six 

 fair geographical species without much character of flower or fruit, 

 though with good habitual marks, and some vegetative characters. 

 The Lake Huron type was low and "few flowered," and had bifid, 

 spiny, overlapping lobes to its leaves, which were tomentose on both 

 sides ; a large subglobose involucre, with appressed lanceolate acu- 

 minate bracts, the spinescent tips of which are presumably spread- 

 ing or reflexed. The heads, in all the western representatives of 

 the aggregate, are ovate rather than subglobose, and have the spin- 

 escent tips of the scales as above described. The pappus of the 

 marginal flowers is only barbellate, while in all the rest in the head 

 it is plumose ; but this it has in common with many other American 

 species. Some of our most striking western plants of this aggregate 

 may be recognized under C. undulatus, as follows : 



Var. Douglasii. Cirnum DouglasiiT). C, Prodr. vi, 643 (1837). 



Stout, probably not tall ; heads rather numerous, not large, short- 

 pedunculate; leaves white beneath, and nearly so above, with a 

 close arachnoid tomentum, deeply pinuatifid, amplexicaul ; heads 

 about 1 inch high, the ovate and ovate-lanceolate bracts closely 

 appressed, with a glutinous spot below the short-spinescent tip ; seg- 

 ments of the corolla shorter than the throat. From Oregon to mid- 

 dle California, in the wooded regions of the coast Range, on rather 

 open ground, hillsides, etc. This is presumably the type of C. 

 Douglasii and very ^wssibly of specific rank. 



Var. megacephalus (Gray, as Cnicus). 



This has few and quite large heads, and a very ample thin foli- 

 age not strongly spinescent. It appears to be a tall plant, and is from 

 the interior of Washington and northern Idaho. Mr. Leiberg's n. 

 654 is doubtless a good type of this. It is also possible that this 

 rather than the preceding may be the true Douglasii. 



