372 THE BOTANICAL EXCHANGE CLUB OF THE BRITISH ISLES. 



Court. Compend. flor. Belgicae manscr." had been validly published 

 already in 1833 in the Magazin d' Horticulture, t. i. . . . and, 

 indeed, on p. 289 with a short French description, and on p. 290, 

 where the characters of the four species, A. minus Schkuhr, majus 

 GmeL, nemorosum nobis, and tomentosum Schkuhr, were set opposite 

 each other, again with brief differential diagnoses. In the Com- 

 pendium itself (1836) the species then appears with a Latin 

 description. The belonging of Lappa maerosperma Wallr. to our 

 species was recently doubted by Moss, who would assign Wall roth's 

 plant rather to A. majus Schkuhr (= A. Lappa L.), yet certainly with 

 injustice, for even if the accuracy of Ray's synonyms as doubtfully 

 cited by Wallroth be questioned, yet Wallroth's diagnosis (" capitulis 

 racemoso-virgatis ") admits of no doubt about the identity of the 

 species. (Conf. also Koernicke I.e., 1864). A. H. Evans (Journ. 

 Rot., Apl. 1913) cites for L. intermedia Lange Fl. Dan. t. 2663, fasc. 

 45, 8, 1844. The three species of Lappa described and figured by 

 Hill (Veg. Syst., 1761) we identify as follows: — L. vulgaris = A. 

 Lappa L. (L. major Gaertner) ; L. arctium = A. tomentosum Miller ; 

 L. minor = A. minus (Hill) Bernh. [To this may be added that in the 

 Ann. Scot. Nat. Hist. 222, 1906, I had come to the same conclusion, and 

 had made A. vulgare (Hill Veg. Syst. iv., 28, 1762) = A. majus Bernh. 

 In Bot. Exch. Club Rep. 195, 1912, I stated that Dr Thellung had 

 shown that A. macrospermum was more recent than A. nemorosum.'] 

 Viola canina L., vice V. ericetorum, Schrad., is used in the sense 

 I have already urged. The authors say: — " In an article which lately 

 appeared 'On the Name Viola canina' (Journ. Bot., Sept. 1911), 

 A. I. Wilmott seeks to adduce proof that by the splitting up of the 

 Linnean collective species V. canina ( = canina auct. + rupestris 

 Schmidt + silvestris Lam. em. Kit. Rchb. + Riviniana Rchb., etc.) 

 the name must in a restricted sense be upheld for V. Riviniana Rchb. 

 (1823) . . . It is meanwhile not difficult to bring forward 

 weighty arguments in favour of the retention of general names in 

 common use for questionable species. At once it follows from Lin- 

 naeus' diagnosis (foliis oblongo-cordatis) and the habitat (in Europae 

 apricis) that the author at least wished the V. canina auct. 

 (ericetorum Schrad.) to be understood under his species, and also the 

 statement 'habitat in pascuis et campis' in the Flora Suecica ( 1 745) and 

 later ed. 2 (1755) which equally is cited after the Hortus Cliffortianus in 

 the first place in the synonymy speaks absolutely for this acceptation 



