Altoiiai'uuj Annual and Biciuiial Habit. 297 



weak, but this treatment has no effect on the period at 

 which the stem w ill be developed. If the seeds are sown 

 in ]\larcli 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 lea\'es, 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 earlv ; 

 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 co\-ered 

 over wnth a sheet every night that threatens to be frosty. 



