— 209 — 



very mixed populations. When 4 characters are considered, we 

 have representatives for respectively 10, 13, 15, 13, and 11 diffe- 

 rent isoreagents in each growing-place, and if the colour of the 

 petals (table II) is also considered, we have from 100 individuals 

 representatives for in E: 27, in F: 21 and inG: 16 different isorea- 

 gents. 27 different isoreagents out of 100 individuals from the 

 same growing-place is a great number! — Therefore those popula- 

 tions are likely to be descended from hybrids between more remote 

 isoreagents. — H, I and K have Æricøfor-character. K is the most 

 pure Zrtco/or-population I have ever found. 



A very interesting fact was observed in the dune -popula- 

 tions I and K. The spur-bearing petal was in most individuals 

 pointed as textfig. i shows. In I 82°/ 0 and in K 90% of the indivi- 

 duals had this point. In dune-populations from Jernen, Sjælland, 

 I have not found this point, nor in any other population except 

 in D, where I found a pointed parv, nib, mnc, ap, alb, nfc. 



In table II the populations E, F, G and H are again put down, 

 but regard is here taken to a fifth character also; in E, F and G 

 the colour of the petals, and in G the branching of the honey 

 streak. As will be seen, tricolor size of petals and labellum can 

 be combined with arvensis colour, and, on the other hand, arvensis 

 size of petals and labellum may be combined with tricolor colour, 

 (as, for instance, in F: gr, lab, mac, cl, alb: 3 individuals, and 

 parv, nib, nmc, ap, v: 3 individuals). 



The branching of the honey-streak was very obvious in the 

 population H (table II), as 73% had branched honey-streak. In 

 comparison with this can be mentioned, that among 500 indivi- 

 duals from 5 growing-places only 33 in all had branched honey- 

 streak. 



The combination parv, lab appears to be difficult to realize, 

 or the individuals with this composition are not capable of living. 

 The number of 4 per 1000 is indeed strangely small. The investi- 

 gation of heredity might solve this question. 



In the spontaneous self-fertilizing forms such as parv, nib, 

 nmc ap, alb and parv, nib, mac ap, alb, which I have cultivated, 

 those 5 characters have proved to be constant, all the descents 

 having the same composition as the mother plant; (for instance, 

 all plants derived from the first one have no dark spot upon the 

 style, while all plants derived from the last one bear a spot). 



These 6 characters contain already a large number of possi- 



Botanisk Tiddsskrift. 37. Bind. 14 



