912 
Journal of Agricultural Research 
Vol. XXIII, No. if 
larly darkened from io a. m. to 2 p. m.; while a third lot was excluded 
from the light only during the early morning period prior to 6.30 a. m. 
Controls were exposed to the full day length, averaging 14 hours and 33 
minutes (sunrise to sunset) for the month of July. The dates of fifst 
flowering in the four series were August 5, August 5, July 27, August 10, 
respectively. It is estimated that in the first three series the reduction 
in sunlight by darkening amounted to approximately 65, 48, and 2% 
per cent, respectively, of the total for the day. It is evident that a 
reduction of 2% per cent in the amount of sunlight when applied in the 
morning was more effective in initiating flowering than a reduction of 
65 per cent applied in the middle of the day. With respect to flowering, 
the remarkable sensitiveness of soybeans to change in length of the daily 
light period is emphasized in the behavior of successive field plantings 
of the Peking; those germinating, for example, on July 20 showed a vege¬ 
tative period 8 to 10 days shorter than those germinating July 5, while 
the change in length of day in the interval of 15 days was only about 17 
minutes, or 2 per cent. It should be noted in this connection that 
Rivera (20), as a result of his investigation, considers that the duration of 
the daily illumination period is more potent than the intensity of the 
light in influencing crop yields. 
To obtain direct experimental evidence on the possible significance of 
the seasonal change in the quality or the composition of sunlight in the 
development of the plant offers considerable difficulty, although this 
phase of the subject is now under investigation. The indirect evidence 
available at this time seems to indicate that this factor is not of great 
importance, at least for the species under study. For example, flowering 
is easily induced in midsummer by shortening the day length through 
cutting off the morning and late afternoon sunlight. The plant in this 
case receives a maximum proportion of sunlight of the shorter wave 
lengths. Flowering also occurs, however, through natural shortening of 
the day length in fall and winter, and in this case the plant receives a 
greatly reduced proportion of the shorter wave lengths. 
As is more fully brought out in the final chapter of this paper, it seems 
likely that the internal water supply of the plant is of fundamental im¬ 
portance in photoperiodism. Nevertheless there is no definite seasonal 
change in rainfall in the eastern United States which would account for 
seasonal periodicity in plant activities. It is necessary to distinguish 
between the internal mechanism of the plant concerned in regulating 
the water content, on the one hand, and the external water supply, on 
the other. In the previous paper (7) detailed data were given to show that 
wide differences in the water content of the soil were without appreciable 
effect in initiating flowering in soybeans. By reference to the experi¬ 
ments with Sagittaria latifolia recorded on page 878 it will be seen that 
growing this plant as an aquatic under a io-hour day delayed flowering 
by only about two weeks as compared with cultures under mesophytic 
conditions. Tlie reduction in the light period, on the other hand, short¬ 
ened the vegetative period by more than two months. 
Temperature undoubtedly is the most important environmental factor 
in relation to the action of the light period on plant growth, and two 
phases of its action need to be taken into account. In the first place, 
the general tendency toward an increased rate of growth and develop¬ 
ment as the temperature increases, within favorable limits, and the 
reverse tendency with decrease of temperature should be to delay or 
even inhibit flowering of typical summer annuals in late summer or fall. 
