On the Movements of Petals 29 



beginning to close in the north house with hght .003, and in the 

 south house with Hght .04, while in the open, with light .15 they 

 were still open. 



These figures all go to show that light is not effective in open- 

 ing and closing flowers, for closing takes place latest in nearly 

 every instance where there is the most light, this being where it 

 is also coldest. The experiment of March 8, 1905, in the shade 

 tent, given on page 27, shows that closing occurs as soon in the 

 darkness as in the light. Morning glories, ready to open, when 

 put in a dark tent at night and kept there all of the following 

 day, behave exactly in opening and closing as they do in normal 

 light with the same temperature. 



It was quite noticeable that when buds ready to open were put 



in too low a temperature (13° or less) they did not open at all 



but assumed the closed position as normally after opening; the 



' temperature was too low for the ordinary processes of life to be 



carried on. 



Experiments to show that humidity of the air is not the cause 

 of opening mid closing. On November 10, 1904, at 5 :oo p.m., 

 three plants, each with buds almost open, were put into different 

 bell jars with light and heat practically the same in all three 

 cases. One was a moist bell jar and one a dry bell jar, with th'e 

 pot wrapped in a rubber cloth, etc. Still others were left out 

 from under the bell jars. At 8:00 a.m. on November 11, 1904, 

 all the buds were open with a temperature of 21° C. On No- 

 vember 14, 1904, at 5 :oo P.M., a similar experiment was per- 

 formed with the same results. 



Experiments with Liniim ttsitatissimum. The flowers of this 

 plant open from 5:00 to 8:00 a.m. and close from 10:00 a.m. 

 to 2 :oo P.M., or, rather, they drop their petals at that time if there 

 is sufficient movement of the air to bring about the process. 

 Otherwise, the petals wither and dry in place. Flax flowers 

 can be caused to drop their petals sooner than normal by in- 

 creasing the temperature. Following are tables showing the 

 results of such experiments : 



219 



