— 380 — 



Characters: 



Size of petals: Stipules: 



parv = petals < the sepals. pinn = pinnate. 



grand ^ petals >- the sepals. palm = palmate. 



Colour of petals: End-lobe of stipules: 



albid = yellowish white. foliac = foliaceous. 



viol, (lut) = blue .or one of the n. foliac = not foliaceous. 

 other tricolor-colovirs. 



by capital letters, which refer to the explanation below. Each 

 of the 16 combinations or isoreagents has a riibric in the table, 

 and the number per cent is stated ofT each habitat. In some 

 cases the number of individuals from a special habitat has been 

 too inconsiderable to be stated per cent, and sometimes no stati- 

 stics have been made. In these cases the Isoreagents noticed 

 in the habitat are indicated by a cross. 



In some of the habitats only arvensis-\ike combinations are 

 found, in others only tricolor-\ike combinations, in others again 

 nearly all combinations are represented. Parv, albid and grand, 

 viol are the two most common combinations, respectively 712 

 and 1130 individuals from 1900. The combinations parv, viol and 

 grand, albid are rare (only 46 and 12 out of 1900). 



Table II next page shows the combinations of some of the 

 characters marking the tricolors from the dunes of Læsø and 

 West-Jutland. Mr. C. A. Jørgensen has been so kind to make 

 up the lists Z, Æ, from Læsø in 1921, as it was of interest for 

 me to know in which combinations the character emarginata was 

 able to occur. (I observed this character in summer 1921 in 

 piants from seeds I had collected on Læsø 1919). For the popu- 

 lations A A and B B only form of stipules, length of calcar and 

 colour of stems are indicated. 



All the Viola in the dunes are grand, lab, non albida {viol) 

 and certainly parvifoliata. The last mentioned character is difficult 

 to compare from different localities, except by cultivation under 

 the same conditions. Most of them are atropiirpurea too. But 

 in other characters all combinations are realized. 



All combinations of characters, I have tried, have proved 

 the same, and there is a multiplicity of possibilities for variations 

 within the collective species. 



