Alternating Annual and Biennial Habit. 297 
weak, but this treatment has no effect on the period at 
which the stem will be developed. If the seeds are sown 
in March in the greenhouse and the seedlings are picked 
out early into pots and planted out in May or June, we 
get vigorous rosettes with abundant leaves, but not a 
single stem in the first year. If the seeds are sown in 
September in the greenhouse, soon after harvesting, the 
rosettes remain weak until winter, but nevertheless de- 
velop a stem in the following spring. By sowing the 
seed in late autumn in the open ground, however, the 
plants will develop only a single pair of leaves above the 
cotyledons and they can be induced to pass through the 
winter without producing their stems in the spring. In 
this case they pass through the whole of the summer 
as rosettes, become extraordinarily vigorous and do not 
develop a flowering stem until after the second winter. 
These experiments show that a definite stimulus is 
necessary for the production of a stem. Under the con- 
ditions of my own experiments it seems to be the winter 
which exerts the stimulus and that it can do so at any 
/ 
age of the plant except the very young stages when only 
the first two leaves are unfolded. But without this stim- 
ulus no stem is formed. 
The experience of beet cultivators goes to show that 
the chief cause of the bolting is the night frosts of the 
spring. Manifestly they exert an effect on the young 
plants similar to that produced by the winter. It is a 
fact generally known that the percentage of bolters is 
high in direct proportion as the seed was sown early; 
crops which have been sown late are sometimes perfectly 
free from this defect. RIMPAU showed that if a small 
section of a field which has been sown early is covered 
over with a sheet every night that threatens to be frosty, 
