039 



name are, that Linnaeus intended Dillenius' Violet, rather than Ger- 

 arde's Violet, under his own application of the specific name "canina." 

 The fact seems to be, that Linnaeus included both species under the 

 single name ; a large form of Dillenius' Violet, and (apparently) a 

 small form of Gerarde's Violet, being preserved as examples of V. ca- 

 nina in his herbarium, three specimens of each. The Linnean Viola 

 canina is thus a group of two species, or perhaps more ; and in sub- 

 dividing this group into its proper and particular species, it would 

 surely have been the better course to restrict the old name of " canina" 

 to Gerarde's Violet, as the species originally intended by it ; instead of 

 dissevering it entirely from the original species, and giving it to that 

 different species which had been distinguished from Gerarde's Violet, 

 by Dillenius, so early as 1724. Tt is reasonable to suppose that, in 

 adopting the old name, Linnaeus intended to continue it to the old 

 species. He may have written the character of his V. canina from 

 a specimen of one of the other species, " lumped" under that name ; 

 but, if so, this was simply an error on his part, in describing the wrong 

 species ofthe group as the true V. canina. 



The question now before us is, are we to adhere to the nomencla- 

 ture of Smith and other English botanists, who have followed the 

 " better course" ? — or, are we to seek future uniformity of nomencla- 

 ture by adopting the new application of the names, as proposed by 

 Fries and many other continental botanists ? There may be no great 

 objection against taking up the name of " sylvatica," instead of " ca- 

 nina ;" for the former would in future be applied more precisely than 

 the latter to Gerarde's Violet. But there is a strong objection to be 

 urged against transferring the name of" canina" to Dillenius' Violet; 

 because in all past English books it means a different species, and will 

 continue to do so in many future books, lists, &c. For instance, 

 henceforth, when we see " Viola canina" in a list of plants, how are 

 we to know whether this name intends Gerarde's Violet or Dillenius' 

 Violet ? — the original V. canina or the substituted V. canina ? 



In the course of the years 1848-9 I have been many times asked 

 for specimens of Violas, and many have been sent to me, with questions 

 about the species and names. And through the obliging aid of Mr. 

 Borrer, this year, in giving me plants of V. lactea and V. Ruppii, 

 which I had not before known in a living state, my collection of the 

 British forms under cultivation is now nearly complete. Some ac- 

 count of the three species may therefore not be useless or out of place 

 here. 



