224 Atkinson, 



small percentage attempted stem development. Two plants came into 

 flower in September. The branches were somewhat more wide spreading 

 than in the normal forms. Two others began stem development in 

 September forming three or four stems each. None of these stems grew 

 erect. All were prostrate and applied closely to the ground. During 

 the latter part of September and October a dense rosette was formed 

 at the tip of each of these prostrate branches, but no roots had formed. 

 The plants with a large central rosette, and several smaller ones on 

 prostrate stems 2 — 3 dm from the central rosette presented a very pe- 

 culiar appearance. 



The longer stem development is postponed during the first season, 

 the more mature and characteristic are the rosettes. In cultures where 

 a percentage only come to stem development during the first season, 

 and these individuals begin stem development at different periods from 

 June to September, great variation in the stem and branching habit, 

 and variation in the size and maturer characteristics of the rosettes is 

 present; the more nearly mature and characteristic the rosettes are in 

 these annual forms, and the more advanced the season so much the more 

 variable may be the habit of the stem stage. 



These variations are epigenetic. They indicate a wide range in 

 the morphological complex, among the individuals of certain species, 

 races or hybrids. The amplitude of this variation is linear, i. e., it ex- 

 tends along the line of the life cycle^which becomes short or long, simple 

 or complex, according to epigenetic conditions. But the lateral varia- 

 tions or fluctuations do not meet nor transgress the limits between the 

 species, races or hybrids. 



In addition to the morphological evidence that there are no inter- 

 grading forms between the hybrids, very strong evidence is furnished 

 by the high state of fertility in hyhrida pycncUa and hyhrida tortuosa, 

 which practically show no variability in this respect, and the almost or 

 quite complete self sterility of hybrida nutella. 



What all of the conditions are which influence this variation in 

 the length of the life cycle, and the lesser or greater degree in the 

 full expression of the maximum characteristics, is difficult to determine. 

 For example, it is difficult to judge the stimulus which determines the 

 beginning of stem development in the annual forms, which in some 

 cases may occur at almost any time during the season. But when stem 

 development begins, it appears to serve as a stimulus which checks the 

 further development of the rosette, so that the rosettes are arrested in 



