THE GENOTYPICAL RESPONSE OF THE PLANT 287 
soon as the shoot reaches the surface of the ground. Usually not 
more than one stem develops in each vegetation period. The plant 
illustrated was collected in 1920. The stem belonging to the shoot ge- 
neration of the preceding year is still attached to the plant as well as 
the one from 1918 (the stump to the right). The leaves are smaller 
and more narrow than in the cultivated plants, etc. In order to follow 
the changes of the plants when moved to the experimental field 15 
specimens were measured and characterized in their natural habitat 
(Vitemölle) in 1920; they were then put into culture and noted in 1921 
and 1922. Table 29 (list no. 28) summarizes the results. The in- 
crease in the growth of the stem is already considerably in 1921. The 
magnitude of the angle of the stem base is about the same in 1920 and 
Fig. 52. H. umbellatum. Cult. mixed population from Torup. 
in 1921, but the terminal part of the stem has turned upwards in 
almost every case. It is now kept at an average angle of less than 
40° with the vertical and this angle is about the same in 1922. The 
change in the width of the leaves is also remarkable; in 1921 the 
have already attained more than twice their original width, while the 
change in 1922 is slight. As to the characters of the leaf margin slight 
changes have taken place, but the characteristics attained in 1921 
remain unchanged in every case in 1922. The reaction-type(TuREsson, 
1922) of this prostrate dune form as well as of the forms of Hofs 
Hallar and Kristineberg discussed above is thus seen to be rather dif- 
ferent in different habitats even if such extreme stages as those typical 
of Centaurea jacea and Succisa pratensis in certain habitats do not be- 
come realized in the case of Hieracium umbellatum. 
The sandfield area at Vitemölle is narrow. Woodlands already 
