MUTATIONS IN VIOLA. 185 



V. FALLACISSIMA. V. Btrnardi, Mackenzie, Man. 135, not 

 of Greene. Low at petalif erous stage, the large foliage on short 

 petioles as in V. Bernardi, but herbage of a light green, the 

 stout petioles quite villous-hirtellous, the peduncles slender and 

 glabrous, bearing the flowers just above the foliage : sepals lan- 

 ceolate, acute, ciliolate : corolla not large, apparently blue rather 

 than purple : foliage whether young or mature more like that of 

 Bernardi than of perpensa, always more or less flabellif orm rather 

 than subcordate in outline, cleft to the middle in the earlier, in 

 the later nearly to the base, at this stage on elongated petioles 

 (plant 8 or 10 inches high) the blade more than 3 inches wide 

 and about as long, the segments broad, obtuse, often coarsely 

 subserrate-toothed, rather densely but shortly hirtellous-ciliate 

 and with sharp partly appressed hairs on the veins as well as 

 now and then in the spaces between them : apetalous flowers of 

 summer yery short- peduncled and almost or quite hypogeous, 

 never straight or upright. 



In dry woods of extreme western Missouri, Jackson Co., col- 

 lected by Mr. Bush and also by Mr. K. Mackenzie, and sent to 

 me rather copiously by both, their specimens with less divided 

 foliage being reported by me as V. Bernardi, the others as V. 

 pedatifida, all too hastily done, on my part ; for not one of the 

 specimens can rightly be referred to either of those species. 

 They all represent a plant somewhat analogous to V. perpensa. 

 yet in character very different. That this has any immediately 

 genetic relation to V. pedaiifida I altogether doubt. All the 

 material known to me is in my own herbarium. 



V. DIGITATA, Pursh. Fl. i. 171 ; probably also V. ranunculi- 

 folia, Juss. in Poir. Encycl. viii. 636. Belated to V. pedata, 

 from which it differs by a cuneate-obovate leaf which is undi- 

 vided and digitately cleft, or merely lobed, or even only toothed. 

 Such a violet, inhabitant of the Atlantic slope of the United 

 States southward, was known to Michaux ; next by Le Conte, 

 who communicated dried specimens to Pursh from Virginia, 

 with also doubtless thb manuscript name V. digitata ; but 

 later Le Conte declined to assign it a name even as a variety, 



Lkaflbts, Vol. 1, pp. 185-200. Feb. 24, 1905. 



