302 THE VIOL^ OF THE COAST SANDHILLS. [October, 



perennial^ or at least biennial^ for I have the withered and 

 bleached stems of the previous year still attached to my speci- 

 men. This root is much branched and penetrates far into the 

 sand, but still produces only one crown, from which spring many 

 stems. The whole plant is thickly set with short hairs. Most 

 even of the lower stipules have their middle lobe crenated and 

 obovate-spathulate ; higher up the middle lobe becomes leaf-like, 

 lanceolate-obovate, toothed, and far broader in proportion than 

 in V. Curtisii. Sepals acuminate and ciliate, as in V. tricolor; 

 their appendages about equalling the corolla-spur. Corolla, in 

 my specimens, altogether of a full yellow colour, the petals 

 rounded and contiguous, much larger than in the Anglesea plant, 

 which however was collected later in the same year. 



I cannot here find any characters by which to separate this 

 plant from V. tricolor, of which Mr. Babington very justly con- 

 siders it a variety. If the root be biennial, or even perennial, 

 the occurrence of biennial roots of V. arvensis in stubbles in the 

 early spring is too well known to allow of much stress being laid 

 upon this point. 



V. tricolor, var. Symei (Baker), is also apparently of longer 

 duration than one year, but its large crenate middle stipules and 

 short corolla-spur equally refer it to V. tricolor. The petals, as 

 shown in the specimen from Mr. Syme, do not exceed in size 

 those of the Portmarnock Viola. It differs however from the 

 latter in being more glabrous, and in having its foliage wider, 

 blunter, and less deeply crenated, even the lateral lobes of the 

 stipules being sometimes widened upwards — spathulate as it were. 

 This I have also observed to take place in the biennial plants of 

 V. tricolor, var. arvensis. 



I find still more reason against identifying V. Symei with V. 

 lutea (as a var.) in the fact that I have gathered upon the sand- 

 hills at Miltown Malbay, in the west of Ireland, a Pansy, which 

 I consider to be most truly V. lutea. In this the flowers are 

 much larger than in V. Symei; the stipules are palmate, with 

 the middle lobe broad, nearly ligulate, bluntish, and quite entire, 

 the middle lobe being less than twice the length of that next ad- 

 jacent ; sepals quite glabrous, with their appendages falling short 

 of the corolla-spur; root far-creeping, not tufted, but, I believe, 

 rhizomatous. 



V. Cwtisii, Forst., is far more distinct-looking than any of the 



