128 • A Miscellany 



open early in the morning. For example, those of the tropical 

 perennial herb, Turnera ulmifolia, open about two hours after 

 dawn, then wither three or four hours later. This occurs in suc- 

 cessive groups of buds even if the shoots are kept in darkness for 

 several days so that they are isolated from the normal day-night 

 changes. However, it is possible to prevent opening by illumination 

 during the night, particularly during the second half of the night, 

 and the anthesis-inhibiting effect of one such illumination lasts 

 for the next three days. Air temperature and relative humidity 

 changes appear to have little effect (Ball, 1933). 



Ball (1936) found similar inhibiting effects of night illumina- 

 tion on morning anthesis in species of Campanula, Geranium, 

 Cist us, and Ipomoea. He determined a crude action spectrum for 

 this phenomenon, using filters, and found that red (6500-7000 A) 

 was the most effective and blue the least effective color. With the 

 advantage of twenty-five years, it is easy to interpret these results 

 as representing the disturbance of a circadian rhythm originally 

 "set" by the light-dark schedule through what is presumably the 

 red, far-red system. However, this work was in a sense before its 

 time, so the (for then) unusual effectiveness of red light attracted 

 little attention. 



A paper by Arnold (1959) on Oenotliera (evening primrose) 

 indicates that endogenous rhythms are also involved here, though 

 relatively susceptible to modification. If the plants receive light 

 from 6 a.m. to 6 p.m. the flowers open at about 6 p.m., as in nature; 

 with an inverse cycle, they open in the morning. Anthesis of a bud 

 that is ready occurs roughly 12 hours after a dark-to-light transition, 

 which thus appears to "set" the timing mechanism, but the timer 

 is easily perturbed by the length of the light period itself. On a 

 schedule of 18 hours light-6 hours darkness anthesis is regularly 

 later, and on 6 hours light-18 hours darkness regularly earlier, 

 than would be predicted by the 12-hour rule. However, it is clear 

 that there is an endogenous component to the timing since anthesis 

 will not follow any arbitrary cycle of light and darkness. The cir- 

 cadian periodicity of anthesis cannot be made into a 12-hour 

 periodicity by schedules of 6 hours light-6 hours darkness, nor into 

 a 48-hour periodicity by alternating 24-hour light and dark periods. 



According to Arnold's investigations the light-sensitive timer of 

 Oenothera anthesis must be localized in the buds. In continuous 



