Vegetative Mutations. 
629 
to obtain any seed (Fig. 137, p. 618), because it flow- 
ered too late in the year. I have cultivated this plant, 
which was described and figured in the first section of 
this volume (p. 92), for many years by vegetative meth- 
ods, because it is perfectly sterile. In the summer of 
1902, however, it suddenly began to produce bud-varia- 
Fig. 143. Cat'pinus Betulus. At a, a partly variegated leaf 
on a tree which had otherwise only green leaves, Hilver- 
sura (1887). 
tions, distributed in considerable numbers over the various 
main stems ; they were apparently independent of one 
another, but seemed to arise in response to the same un- 
known external influences. The flowerheads of the ata- 
vistic branches were of the normal structure of the ordi- 
nary Dahlias, double, and with carmine red florets, at the 
tip of each of which \vas a white spot. From this we 
