VIOLA TRICOLOR VAR. SABULOSA AND ALLIED FORMS 343 



ovatis elongatisve, sepalis angusto-lanceolatis corolla vix brevi- 

 oribus. — in arenosis maritimis Belgii et Galliae." Here, De Candolle 

 says nothing about the colour of the flowers, nor of the differential 

 characters of the stipules and relative length of leaves and 

 peduncles, as shown in the description of the two forms given 

 further on. Nor does he refer to the sandhill Violet mentioned 

 by German authors a few years previously. There is no reference 

 to these German opuscula in the " Bibliotheca Botanica " which 

 prefaces (and supplements) his Sijstema Naturah (1818-1821). 



In Journ. Bot. 1909, suppl. p. 13, Dr. E. Drabble defines a group 

 CurtisiecB : — " Perennial plants of maritime sandhills, with sub- 

 terranean slender branching perennial stems. Flower with petals 

 longer than the sepals, and with long petaline spur." This group 

 I consider equal to and identical with Syme's V. tricolor subsp. 

 Curtisii — -but preferring Jordan's term of " grex " for " subspecies." 

 In the interests both of convenience and of scientific accuracy, it 

 will be best to conserve the three varietal names of Curtisii, Pes- 

 neaui, and sabulosa as constituting " Grex viaritima " { = subsp. 

 Curtisii). Schweigger's earliest name of maritima is thus kept 

 up to include distinct forms and varieties of the sandhill plant 

 which he defined only in the aggregate, and known to him only 

 from the German coast. 



Walking along the dunes from Knocke to the Dutch frontier, I 

 found var. sabulosa fairly plentiful, and growing under two forms, 

 which I describe below as f. discolor and f. concolor, the latter 

 readily distinguished from var. Pesneaui, which, though it has 

 similar blue-violet petals, is quite hairy and csespitose as compared 

 with var. sabulosa. Last year I observed it similarly along the 

 dunes in walking from Ostend to the French frontier, especially 

 near Nieuport. 



In Journ. Bot. 1909, suppl. p. 14, Dr. E. Drabble aptly 

 remarks : — " This plant differs from the other members of the 

 Curtisii section in having long and very narrow leaves. Some of 

 our British Pansies seem to be referable to this species, but I 

 have not seen any with leaves so long and narrow as certain 

 specimens from the West of France." 



Going through the material in the British Herbarium of Herb. 

 Mus. Brit., the nearest Britisli specimens to var. sabulosa which 

 I have seen are, one from Ross Links, Northumberland {H.E.Fox, 

 1884), and the other from Pembrey Burrows, Caermarthenshire 

 {E. S. Marshall, 1899). In the first, the stem and leaves are 

 distinctly hispid-hairy, the bracteoles are too high up on the 

 peduncle, and the upper leaves are not narrow enough. In the 

 second, the stem and leaves are also hispid-hairy, there are too 

 many leaves on the lower part of the stem, and the leaves them- 

 selves are not long enough to match the Belgian plant. 



Many notes were made from individual plants growing in situ; 

 and in several instances the long root was dug out of the sand 

 down to its lowest fibres. It is certainly perennial, like the 

 British forms of the subspecies Curtisii ; not biennial, as stated 

 by Frencii authors, or annual, as stated in Belgian plant-lists. 



