191 



Echinncea nnrpiatifolia DC. " By Michigan Southern and Lake Shore Rail- 

 road, Durham, Ind. In a prairie. July 4, 1892." (E. J. Hill.) So far as I am 

 able to find, this is the only record for this species in the State. The form 

 has evidently entered our flora from the west, its recorded range being "Plains 

 from 111. and Wise, south westward." li is easily distinguishable from E. pur- 

 purea Moench. and should be looked for carefully in the western counties of the 

 State. 



Artemisia Canadensis Mich x. "Shores of Lake Michigan, Lake Co., Sept. 4, 

 1898." (E. J. Hill.^ This northern form has its only recorded station for Indiana 

 in the above reference. Its range is "Northern New Eng. to the great lakes, 

 Minn., and northward." It is closely allied to A. caudata Michx., which also has 

 its .sole Indiana station in Lake Co. No specimen of this latter form, however, 

 has as yet been obtained by the Survey. A. caudata having a range "Mich, to 

 Minn., and southward," should be found, at least, in the northern counties of the 

 State. Both species are separated from the other Artemisias by their dissected 

 leaves and should be readily recognized. 



Cnicus Piichrri Torr. "Sandy shores of Lake Michigan, Pine Station, Ind., 

 June 21, 1891." fE. J. Hill.) This well-marked species has this as its only sta- 

 tion in the State, so far as the records indicate. Its range, " Sandy shores of 

 Lakes Michigan, Huron and Superior," would indicate but a slight probability of 

 any material increase in its distribution. It would probably be found in Laporte 

 County in the region of Michigan City, if careful search were made. With its 

 cream-colored flowers and white woolly covering it is an extremely attractive 

 form and could scarcely be mistaken for any other species of the genus. 



Cnicus pumilus Torr. "Pine l)arrens. Lake County, Ind., July 4, 1891." 

 (E. J. Hill.) This form is labeled Cnicus Hillii W. M. Canby. lam unable, 

 however, to see any reason why the form should not be referred to C. pumilus 

 Torr., and in the absence of Mr. Canby's original description I have so referred 

 the specimen sent to the Survey. Certain variations from the type seem to me 

 easily referable to geographical causes, and not of sufficient importance to neces- 

 sitate the establishment of a new species. The range of the plant, "Dry fields, 

 N. Eng., near the coast, to Penn.", seems to me to furnish the only argument 

 against the reference. It is possible that more abundant material may lead to a 

 difTerent conclusion. The reported occurrence of Cnicus pumilus in Dearborn 

 County (S. H. Collins) is not authenticated by specimens, and is in all probabil- 

 ity an error in determination. The extension of the range of a coast plant to 

 the Great Lake* could be easily accounted for, but its extension to Dearborn 

 County without intervening station* would be difficult of explanatinn. 



