222 THE JOURNAL OF BOTANY 



specimens from Schultz Herb. Normale no. 1019 (from granitic 

 escarpments between Minister and Gerarclmer, in the Vosges), 

 which have been referred to this variety. Although, as just stated, 

 it is closely allied, there are several differences, and I venture to 

 describe it as a subvariety : — 



Var. AMCENA subvar. insignis. Perennial. Stem rather short or 

 somewhat elongate (5-15 cm.). Lamina of lowest leaves orbicular, 

 several times longer than the petiole, base rounded or subcuneate, 

 margin crenate-serrate ; lamina of upper leaves ovate-oblong or 

 oblong, differing from our specimens of V. lutea var. c/randiflora 

 in several points — in the obtuse apex, longer petiole, and broader 

 lamina ; the petioles and lamina are more or less hairy. Stipules 

 palmately pinnatifid ; terminal lobe rather larger than the others. 

 Peduncle 5-6 cm. long. Bracteoles placed below the curvature. 

 Flower about the same size as that of var. n r and i flora, but upper 

 petals narrower. Sepals subacuminate. Upper petals divergent, 

 narrow obovate, + 1*8 cm. long, 8-9 mm. broad at broadest point — 

 lateral petals spreading, lower petal rather broader than long, 

 + 1'3 cm. long, + 1'7 cm. broad at broadest point. Spur just 

 longer than the appendages of the calyx, not so long as in var. 

 c/randiflora. Capsule sliorter than sepals.. 



Hab. South side of Craig Caillaich above Fnilarig, in Breadal- 

 bane ; and rocks, somewhat moist, at very considerable height on 

 Ben Lawers, frequent, Aug. 1794, Robt. Brown, in Herb. Mus. Brit. 

 Cliffs of Ben Lawers, G. C. Driice, Aug. 1888, Herb. G. C. Druce. 



It is larger-flowered than typical var. amcena, and has, as has 

 been stated, a stem sometimes 15 cm. long. The large size of the 

 flowers (which are purplish) makes this a very striking plant. 



I am unable to follow Messrs. Rouy and Foucaud in their de- 

 scription of V. lutea a nnf/uiculata, for which they quote the following 

 synonymy : — 



" V. grandiflora Huds. Fl. Angl. ed. 2, p. 380; V. lutea var. 

 yrandiflora Koch, Synopsis, ed. 2, p. 95 ; G. et G. Fl. Fr. i. p. 185." 



F. grandiflora Hudson, Fl. Angl. ed. 2, p. 380, has generally 

 been considered synonymous with V. lutea Hudson, ed. 1, p. 331 '■' ; 

 but it should be noted that Hudson under V. grandiflora quotes 

 Viola caule triquetro simplici foliis oblofigiusculis stipulis pinnatijidis of 

 Linnaeus (Mantissa, p. 120), and Viola montana lutea grandiflora 

 Bauhin, Pinax, 200, which are both placed by Linnaeus (Mant. I.e.) 

 under his grandiflora. 



V. lutea Hudson a unguiculata Rouy & Foucaud is for the most 

 part then synonymous with V. lutea Huds. var. grandiflora Koch, 

 but is confused by these authors in their synonymy with V. lutea 

 Hudson, Fl. Angl. ed. 1, p. 331. V. B. Wittrock, in his "Viola 

 Studien," i. (in Acta Horti Bergiani, Band 2, No. 1, p. 96) describes 

 and figures V. grandiflora Lin. Vill., but his fig. 107 on tab. vii. 

 shows considerable difference in the character of the stipules from 

 those organs as figured by Villars. 



* Confer Smith, English Flora, p. 307 ; and Koch, Synopsis, ed. ii. p. 95 

 (1843). 



