Time Relations and Endogenous Rhythms • 49 



crucial differences (evidence for the second, or middle, maximum) 

 were very small, and reached opposite conclusions on other grounds 

 (see below); but very clear data confirming Carr's results were later 

 published by Melchers (1956). Meanwhile, Hussey (1954) had shown 

 that the LDP Anagallis arvensis grown in 72-hour cycles with long 

 dark periods showed only two maxima for the promotion of flower- 

 ing by light-breaks instead of the three that would correspond to 

 Carr's results. With Hyoscyamus, however, Clauss and Rau (1956) 

 were able to show three optima in similar experiments, thus sup- 

 porting Carr and Biinning. The quantitative LDP Arabidopsis 

 thaliana was studied twice, with ambiguous results each time (Hussey, 

 1954; Clauss and Rau, 1956). The SDP Coleus blumei x frederici 

 disagreed with all others, since the time for maximum light-break 

 inhibition (72-hour cycle) was in the middle of the long dark period, 

 with no sign of three or even two optima (Kribben, 1955). 



Other work besides that on 72-hour cycles suggests Carr's 

 quoted conclusion may have been hasty. Wareing (1954) voiced 

 strong opposition to the idea that endogenous alternation of photo- 

 phile and scotophile phases determines the action of light-breaks. 

 He presented experiments with Biloxi soybeans grown on 9 hours 

 light-39 hours darkness (48-hour cycles), or on 9 hours light-51 

 hours darkness (60-hour cycles), testing the effects of light-breaks 

 at various times during the long dark periods. In both cycles light- 

 breaks about 6 to 8 hours before or after the main light periods 

 were maximally inhibitory, whereas they promoted flowering in 

 the middle of the dark periods. Since the dark periods used in the 

 two cycles differed by 12 hours, one would not expect these results 

 if the inhibitory action of light-breaks was due to a more or less 

 unchanged circadian rhythm. One would expect them, however, if 

 light-breaks interact with the main photoperiod when it is close 

 enough, thus providing a total photoperiod that exceeds the "limit- 

 ing value" for soybean flowering (see Chapter Two). Further evi- 

 dence for this view was that in cycles totaling 48 hours, light-breaks 

 given either 3 or 6 hours before the main light period were inhibi- 

 tory when the latter was 9 hours long, whereas only a light-break 

 6 hours before was effective with a 6-hour main photoperiod. 



Wareing also reported experiments with Xanthium in which 

 light-breaks toward the end of a long dark period were not inhibi- 

 tory. Since this plant, unlike soybeans, has no "limiting photo- 



