292 THE JOURNAL OF BOTANY 



cited by Linnaeus, particularly the descriptions by Bauhin, Mori- 

 son, and Haller. 



The views of Elias Fries, a great student of violets, will be 

 now considered. In his Novitice Flora Suecicce (1828), p. 2, Fries 

 uses the name Viola canma to include all the wood violets, heath 

 violets, and sand violets (F. arenaria DC, V.rupestris Schm., &c.). 

 This is done intentionally, as is witnessed by the synonymy cited, 

 and also by his remarks (in Mant. iii. p. 118) in the first paragraph 

 of his discussion of V. canina. A translation of his remarks is 

 here given : — " The Linnean plant is manifestly an aggregate, for 

 though it may now be possible from the words of his definition 

 to relate the name successfully to a definite plant, yet for the sake 

 of avoiding controversy, Linnaeus ought scarcely to be cited as 

 the author of the name, for there is no doubt that Linnaeus very 

 often had V. arenaria and V. silvatica before his eyes, which are 

 the forms by far the most copiously and frequently met with in 

 the neighbourhood of Upsala. It is certain that V. silvatica alone 

 is present under this name in the Linnean Herbarium, wherefore 

 De Candolle and many others have taken V. silvatica for the real 

 V. canina, and Smith distinguished V. canina Rect. under the 

 name V. flavicornis, calling V. sylvatica V. canina. If these be 

 all joined together, Linnaeus is the author of the species." 



We have already shown that some of these statements are in- 

 correct, and that the Linnean plant is definitely a wood violet. 

 Again, Fries's position with regard to nomenclature is well known 

 to-day, but it is now fairly generally agreed that it is desirable to 

 retain the Linnean name '•' wherever practicable. With regard to 

 the plants. Fries agrees with Reichenbach, having used the latter's 

 exsiccata for help in identification {Nov. I. c). It seemed 

 necessary to state Fries's position because Babington refers to it 

 when stating his own opinion. 



Babington, in his Flora of Cambridgeshire, p. 302 (1860), 

 Appendix iii., gives reasons for adopting Eeichenbach's position. 

 He says : " Ray remarks of his plant, named V. canina sylvestris, 

 that it grows ' ad sepes et dumetis passim. Habetur et in palustri- 

 bus frequens, nisi forte ea sit distincta species ' {CaJ,. Angl. ed. i. 

 317). The former is therefore our V. sylvatica,} the latter V. canina. 

 Linnasus is supposed by some botanists to have derived his plant 

 from the books of Ray and Gerard ; but he quotes neither 

 of them in the Hortus Cliffortianus when founding the species. 

 His character, ' V. foliis cordatis oblongis, pedunculis fere radi- 

 catis,' will not apply to our V. sylvatica, nor does the cut in 



* If the Linnean specific names are not to be retained, it is only logical to 

 treat the genera in the same manner. The genus SalicoDiia Linn., as was 

 pointed out to me by Dr. Moss, has now become the tribe Salicorni<c Du Mortier, 

 including (in Dalla Torre, Gen. Siphonog.) eleven genera and thirty-three 

 si)ecies, and yet the Linnean name Salicornia, with only nine species, is univer- 

 sally retained for one of these. 



t Viola sylvatica Bab. Manual, ed. 5-8, is an aggregate of V. ST/lvestris Kit. 

 and V. Riviiiiana Echb. V. silvatica Fr. = V. sylvestris alone, V. Eiviniana 

 being called V. silvatica var. viacrantlia (Fries, Mant. iii. p. 121). 



