— 397 — 



construct a system of Isoreagents from this list according to the 

 special investigations for whicli he will use them. 



The splitt ing up of Linnean species in new »species« has fre- 

 quently been based upon accidental observations of deviating in- 

 dividuals which are described, instead of trying to make a per- 

 spicuous classification of the types within the species based upon 

 studies in the extent and the cause of the variation and other 

 questions in connection herewith. In many cases this result is due 

 to the faet that these systematists either do not accept the conse- 

 quences of, or are not fully famihar with all the results of modern 

 biology. The systematist's investigations are so difficult, that it 

 is needful that he uses all the resources at his disposal. 

 At the investigation of the small systematic units within the single 

 Linnean species he cannot be content with employing the old me- 

 thods alone, which suited the classification of the Linnean species 

 themselves so long ago. 



VI. Must Viola tricolor and arvensis be regarded as two 



distinct Species? 



During the last 200 years these two systematic units alter- 

 nately have been regarded as two distinct species and as two minor 

 units belonging to the same species* 



As early as in 1745 Alb. Haller distinguished two species 

 (Haller 1745). In Historia Plantarum Helvetiae (1769) 

 he distinguishes the two species in the following way: 



n. 568 {V. tricolor Riv.): "flore calyce duplo longiore." 



n. 569 (V. bicolor Riv.): "flore calyce paulo majori." 



LiNNÉ regarded them as one species (Species Plantarum 

 1753). 



Murray (Prodr. Stirp. Gottingensium 1770) separates 

 them under the names V. tricolor L. and V. arvensis Murr. (syn: 

 V. bicolor arvensis C. Bauh.) and for description he refers to 

 Haller's forenamed description of them both. — Most frequently 

 they have been looked upon as two species. 



I regard them as forming two species. If we assume the 

 criterion of hybridization as the only criterion in all cases, they 

 would not be so. They cross and produce fertile offspring. And 

 moreover, there are smooth transitions between them. But then 

 all the Melanium section would constitute one species. In Viola 



