ECOLOGY AND THE FLOWERING PROCESS 37 



somewhat more accurately in summer (or winter) than during the 

 spring and fall months. For example, during spring or fall a plant 

 at 40° latitude must measure the diurnal cycle within about 26 min if 

 it is to flower within a 10-day period encompassing the "proper" 

 time. If the night length is measured, 26 min is about 4.3% of a 

 10-hr night coming at the beginning of August or 3.6% of a 12-hr 

 night near the end of September. If, however, the response occurs 

 about July 9, the 9-hr night will have to be measured with an accuracy 

 of 10 min to insure flowering within the 10-day period, requiring an 

 accuracy of about 2 %. 



Another problem which arises in trying to understand the responses 

 of plants in their natural situations by using data obtained in our 

 experimental studies, is that we have yet to study in any detail the 

 ability of a plant to measure time as the day-length is changing. It 

 is clear, however, that many plants measure time more accurately 

 when they receive a number of repeated cycles than when they receive 

 only one or a few. With cocklebur this is evident when the effects of 

 only 5 cycles are compared with those of a single cycle, as in Fig. 3-7. 

 With a single cycle the plants of a population varied as much as an 

 hour among themselves in their measurement of the critical night. 

 With five cycles this error is reduced to about 20 min. 



Response to day-length implies directly a measurement of time. 



Figure 3-5 

 Day-length and time of sunrise at various latitudes as a function of time 

 during the year. Only the dotted curve shows true day-length (60° lati- 

 tude), the others show sunrise times. The difference is slight and can 

 probably be ignored. The latitudes shown pass near the following 

 geographic locations: 



30° latitude, New Orleans, Cairo; 



40° latitude, Denver, Philadelphia, Madrid, the "heel" of Italy; 



50° latitude, Vancouver, Winnipeg, the southern tip of England, 



Frankfurt; 

 60° latitude, Anchorage, southern tip of Greenland, Oslo, 

 Helsinki. 

 Data from Astronomisch-Geoddtisches Jahrbuch, 1954. G. Braim, 



Karlsruhe. 



Figure 3-6 



Rate of change in day-length at two latitudes as a function of time during 



the year. Data from Astronomisch-Geoddtisches Jahrbuch, 1954, 



G. Braun, Karlsruhe. 



