260 THE JOURNAL OF BOTANY 



synonyms belonging to V. montana L. and also " V. monfcana, 

 Linn. Sj^ec. 2, jj. 1325, ex herb. & syn., speciatim Royen. ! Dec. 

 Prodr. 1, p. 299, a ! (3 ! — omniumque antiquior. speciatim Roth. 

 Germ. 2, p. 270!! minim, ejus V. persicaef," but lower down adds 

 " nomen vero ad V. montanam alteram Fl. Suec. pertinot." Of tliis 

 latter (as V. canina y montana) he says " V. montana Linn. Slice, 

 n. 967 ex loco. Wahl.l Lajjp. n. 399, Suec. n. 967." 



Burnat and Briquet op. cit. state the case from a modern 

 point of view. They consider that F. monta?ia L. Sp. PI. was an 

 aggregate of V. stricta, V. stagnina and V. elatior, and that the 

 name can therefore be used for any one of these three. They also 

 consider that in the Flora Suecica Linnaeus used the name for 

 V. stricta Fr. alone, and remark " Linn6 . . . cite encore 

 quelques synonyms douteux, mais il ne fait plus mention de 

 Morison dont la figure se rapporte bien au V. elatior Fries." It has, 

 however, been sufficiently pointed out how incorrect it is to call 

 Linnfeus's synonyms " douteux," and it remains to be shown that 

 V. montana L. is an aggregate. Even if the original descriptions 

 are somewhat vague, they were quite definitely and consistently 

 understood by Ray, Parkinson, Plukenet, Tournefort, Morison, 

 etc., and why not also Linnaeus? If anyone nowadays records a 

 find of ''Galium saxatile Linn." or of '' Callitriche verna Linn." it 

 would be considered quite a definite statement, in spite of the fact 

 that the original references have often been considered " douteux." 

 But even if the premises were correct, this using of local floras to 

 precise names " ex loco " is illogical. Obviously the author of a 

 local flora is in a sense only dealing with those forms of the species 

 which grow in his area, but he cannot in any sense be regarded as 

 restricting the name to those forms. To take an example, Hudson 

 (1762) Fl. Angl. p. 209 has (see Journ. Bot. 1907, p. 435) been 

 regarded as restricting the name Adonis annua to the only British 

 species, viz. A. autu^nnalis. This is not so. All that Hudson 

 means is "The British Adonis belong to A. an7iua Linn., other 

 varieties of which occur outside Britain." The name A. annua 

 undoubtedly must, according to the international rules, be 

 restricted to one species of Adonis, and this is done by Miller 

 {Gard. Diet. 1768) who keeps up his two species as A. anmia and 

 A. cestivalis. But the plant must be cited as A. annua L. emend. 

 Mill., and not A. annua L. emend. Huds., Miller being the first to 

 use the name with the new connotation. Names in local floras, 

 unless the contrary is definitely stated, are to be regarded as 

 identifications, the author merely referring his plant to a known 

 species. To regard them as restrictions is to fall into the " fallacy 

 of the undistrilDuted predicate," due to the neglect of the very first 

 theorem of logic. Linnaeus identifies the Lapland Viola with V. 

 montana Sp. PI. (1753) and also of ed. 2 (1763), in which Morison 

 is still quoted, the account of the species being unaltered. If his 

 identification is incorrect V. elatior must be called " V. monta^ia 

 Linn, excluding all references to the Swedish plant." Borbas has 

 rightly done this in Hallier's edition of Koch's Synopsis. 



But in any case Fries only assumed that F. elatior did not grow 



