VIOLA 79 



distributed through the county ; it attains its maximum of frequency 

 in light sandy soils, and is also common in the flinty arable fields on 

 the chalk. 



In our British lists V. tricolor is separated into two species, the V. 

 eu-tricolor of Syme and the V. arvensis of Murray, Prod. Stirp. Gott. 73 

 (1770) =var. ARVENSIS, Linn. The former is characterized by the 

 petals being longer than the sepals, while in the latter they are 

 usually shorter than the sepals, but they may be also of the same 

 length ; the capsule in arvensis is said to be more globular. I have 

 departed from the practice of some of our British botanists by com- 

 bining the two plants under one head. I find in fact such a range of 

 variability in the species that it would be necessary to give specific 

 rank to many of its forms, if such artificial and variable characters 

 as the length of the petals were chosen for the definition of species. 

 Continental botanists have described a considerable number of micro- 

 species, and it would appear that some of them come true from seed. 



Var. BELLA, Gren. & Godr. Fl. Fr. i. 184, which is a large-flowered 

 plant with the petals usually pale yellow. It is rather frequent in 

 arable fields on the high ground near Streatley and Yattendon, and 

 at Bradfield and Beenham. Probably it is the var. Curtisii of Pamplin 

 in Phyt. v. 154 (1854). 



Var. BicoLOR, Hoffm. ex Reichb. Fl. Germ. Exc. ii. 710- Specimens 

 agreeing with plants thus named in Baenitz' Herb. Europaeum, from 

 Prussia, have large flowers, the upper petals being more or less 

 purple coloured. Such plants have been seen near Upper Basildon. 

 Both of these forms belong to V. tricolor proper. 



In Journ.Bot. (1870), 223, Mr. Britten referred the cornfield pansy of 

 South Bucks to F. Paillouxii [sic Paillouxi, Jord. Obs. sur PI. Crit. ii. 

 (1846), 36], and in his Contributions (p. 45), said he had seen it at 

 Cookham and V^argrave, and suspected this to be Pamplin's var. 

 Curtisii. I cannot reconcile this statement with that expressed by 

 Mr. Britten in i?ep. of Bot. Exch. Club, 1882, where he writes, ' we seem 

 to have all imaginable forms of (F. tricolor) here except F. Paillouxii, 

 which I cannot find.' 



Of the forms with small petals which would be grouped under 

 F. arvensis, we have var. segetalis (Jord. Obs. sur Plantes Crit. 11. 

 (1846) 12, as a species), which is a form with very small petals and of 

 somewhat lax habit, so that the internodes are longer than in 

 ordinary arvensis. 



Var. PUBESCENs, Willd., has been issued by Wirtgen in the PI. 



Select. Fl. Rhen. No. 400 ; it is a more pubescent form with small petals. 



Var. MENTiTA (Jord. in Billot. Annot. loi, as a species) has been 



gathered near Bucklebury. These are intermediate between the two 



extremes. Over the Isis and Ock districts the small-flowered forms 



