EFFECT OF LIGHT ON PLANTS GARNER AND ALLARD. 573 



neither advanced nor delayed by a single day. Similarly, it was 

 observed that the Maryland Mammoth tobacco is not affected by 

 shading so far as concerns date of flowering. We conclude, there- 

 fore, that change in the intensity of the sunlight as the season ad- 

 vances is not the factor which hastens flowering and fruiting in the 

 Biloxi soy bean and the Mammoth tobacco. As a matter of fact, 

 the intensity of the sunlight varies each day from zero just before 

 the beginning of dawn to a maximum at noon, after which it again 

 declines to zero at the end of twilight in the evening. Moreover, 

 periods of relatively dark, cloudy weather of variable but consider- 

 able duration occur at irregular intervals during the growing sea- 

 son in many sections. It is hardly to be expected, then, that seasonal 

 changes in light intensity would be a factor of importance in such 

 features of plant development as the sharply defined annual perio- 

 dicity in time of flowering and fruiting shown by soy bean and to- 

 bacco, as well as by most other plants when grown outside the 

 Tropics. 



EFFECT OF SHORTENING THE DAILY ILLUMINATION PERIOD ON 

 THE DEVELOPMENT OF SOT BEAN AND TOBACCO. 



Having seen that changes in temperature and in the intensity of 

 the sunlight do not hasten flowering and fruiting of soy bean and 

 tobacco as the season advances, we next turn to another feature of 

 periodicity related to change of season, namely, the relative length of 

 day and night. In the latitude of Washington, D. C, practically 

 15 hours elapse between sunrise and sunset during the longest days 

 of the year, which occur, of course, during the latter part of June, 

 while during the latter part of December there are only about nine 

 and one-half hours between sunrise and sunset. Beginning with the 

 first part of July the length of day decreases by slightly less than one 

 minute daily. This rate of decrease becomes steadily larger until the 

 end of September, when it is considerably more than two minutes 

 per day, while from that time till the winter solstice the rate of 

 decrease steadily declines. The change in the length of the day at 

 Washington during the principal growing season for plants is shown 

 in figure 1. 



To ascertain whether the seasonal change in the length of the day 

 is in any way responsible for the peculiar behavior of soy bean and 

 tobacco under discussion it was decided late in the summer of 1918 

 to make a simple experiment with these plants. A small, A^entilated, 

 light-proof chamber was constructed into which boxes containing 

 soy bean and tobacco plants could be placed for a time each day. 

 so as to reduce the number of hours of sunlight received by the 

 plants. This dark chamber is shown in plate 3. In practice the 

 cultures of soy bean and tobacco were placed in the dark chamber 



