Lilium Liliaceae 313 



spreading of the perianth is a note often omitted, and herbarium specimens 

 do not always clearly show this character. The map shows the location of 

 the specimens that I have seen. Birkbeck passed through Indiana in 1817 

 and on page 112 of his "Notes on a journey in America from Virginia to 

 the Illinois Territory" he says : "The road from Sholt's tavern to this place 

 [from thirty-six miles east of Vincennes to Vincennes] is partly across 

 barrens, that is, land of middling quality, thinly set with timber, or covered 

 with long grass, and shrubby undergrowth; generally level and dry, and 

 gaudy with marigolds, sunflowers, martagon lilies, and many other bril- 

 liant flowers." We have no way of determining the species of lily Birkbeck 

 saw, but I think it was this species, although this is a mere guess. 



From Lilium michiganense this species can be distinguished by several 

 characters in addition to those given in the key. It is about half again as 

 high; the leaves are more numerous in at least a few of the whorls, and 

 all the leaves are longer and conspicuously long taper-pointed at both ends ; 

 the more numerous flowers, 3-15, are in umbels or pyramidal racemes, 

 while the flowers in the other species are strictly in umbels, generally num- 

 bering 1-5 flowers. We have had both species moved from the wild and in 

 cultivation for many years and the greatest number of flowers of the first 

 has been 21 while of the second I have no record; but, as I recall, the 

 number ranges mostly from 3 to 5 in vigorous plants, and these are always 

 in umbels. 



N. B., Ont. to Minn., southw. to Va. and Mo. 



3. Lilium canadense L. f. rubrum Britt. (Bull. Torrey Bot. Club 17: 

 125. 1890.) Canada Lily. Map 644. This lily has been confused with the 

 next, and possibly all, or nearly all reports for it should be referred to the 

 next species. All of my specimens are from rocky, wooded slopes and were 

 associated with Vaccinium vacillans and Asclepias tuberosa. This species 

 is now considered to be Alleghanian and its distribution is not known 

 because of its confusion with the next species. It is known to occur near 

 Lawton in Kalamazoo County, Michigan (Nieuwland). 



4. Lilium michiganense Farwell. (Bull. Torrey Bot. Club 42: 352-354. 

 1915.) Map 645. Infrequent throughout the northern part of the state and 

 probably rare in the southern part or even absent from the southeastern 

 part. It has been confused with the preceding species and our knowledge 

 of its distribution and habitat can be now ascertained only from field 

 studies or from existing specimens. Almost all of the reports for lilies in 

 Indiana must be ignored on account of the recently acquired knowledge 

 of the genus. 



Lilium michiganense grows in moist prairie habitats, in mucky soil about 

 lakes and in low woods, and in moist, black soil along roadsides and rail- 

 roads. Locally it may be common over a small area. When once established 

 it is very persistent, competing successfully with blue grass sod. I have 

 known it to be a common plant for possibly 25 years in black, moist soil 

 along the railroad through the old prairie north of Poneto, Wells County. 

 It is to be noted with this species, as with the others, that the available 



