VIOLACE^. 



27 



This plant differs remarkably from the two preceding forms 

 in branching- below ground instead of above it, as they do. Tiie 

 lowest stipules are also shorter, with fewer lateral divisions. The 

 terminal lobe varies in being entire or crenate ; but as this occurs 

 also in unmistakable V. eu-tricolor and V. arvensis, it cannot be 

 deemed a character of any importance. The stems are shorter, more 

 decum])ent, and more decidedly rough with short hairs, than in the 

 two preceding plants. The flowers have spreading petals, much larger 

 than tliose of V. arvensis. In the Mullaghmore plant the flower 

 measures 1 inch from the top to the extremity of the lowest petal. 

 Iliis form has been named V. Symei by Mr. Baker, and has the 

 stipules much larger than those of the otlier plants here associated 

 under V. Curtisii ; but its mode of growth and form of the lowest 

 stipules are precisely similar to that of the New Brighton and 

 Port Marnock forms, which appear to me to be inseparable from the 

 oi'iginal V. Curtisii, i.e. that which is found at Braunton Burrows. 

 Mr. J. G. Baker (who has paid much attention to the British 

 Pansies, and to whom I am much indebted for copious notes 

 respecting them, with which at my request he has kindly furnished 

 me) has sent me a Pansy collected by himself near the Spital of 

 Glenshee, Perthshire, which he believes to be V. lepida (Jorcl.), and 

 which lie is inclined to refer to Curtisii rather than V. eu-tricolor 

 or V. lutea. This is the plant mentioned in the " Journal of 

 Botany" for January, 18G3, p. 11, where it is mentioned as having 

 the " habit of growth of V. (eu-) tricolor, but yet apparently with a 

 perennial root, and growing in a station suitable for V. lutea, in 

 a meadow near the banks of a stream." Judging from the speci- 

 men, it appears to me to be referable to V. eu-tricolor, of which it has 

 the habit and all the characters except the annual root; it is rather 

 more liairy than usual, but the long erect stem branching al)ove 

 ground appears to me to give conclusive evidence as to which of the 

 Ibi-ms it belonjxs. 



Sea Pan SI/. 



o 



Sub-Species IV. — Viola lutea. ihuls. 



Plate CLXXXI. 



neich. lo. Fl. Germ, et Helv. Vol. III. Viol. Tab. XXIII. Fig. 4.519. 

 V. svulitica, Wild. 



Pi.ootstock branched, with the branches very slender, not tufted, 

 producing short or rarely elongated ascending stems and long 

 slender subterranean stolons. Stipules of the lowest leaves 

 digitate-pinnatifid. Petals generally longer than the sepals, spread- 

 ing. Capsule oblong-ovoid, 3-sided. 



In grassy places. Common in hilly and mountainous districts. 

 It occurs in Wales ; Nottinghamshire and Derbyshire appearing 



