sept. i5 ,1925 Response of Plants to Length of Day and Night 
561 
Beginning on the same date, another test was made in which the 
treatments just described were reversed, the upper portion of the 
plant being exposed to the full period of daylight, while the lower 
portion was exposed for only 10 hours daily (pi. 3 a, d). Flower buds 
appeared in lower portion of plant July 3, and the first blossom 
opened July 17. No flower buds appeared on upper part of plant 
until October 1 after it had been transferred to the greenhouse. 
Here, again, the vegetative and the flowering zones of the axis were 
sharply delimited by the two light treatments. 
The last experiment, in conjunction with that first described, 
shows that flowering has been readily induced on the lower portion 
of the stem by exposure to a short daily light period, while flowering 
in the upper part of the stem was inhibited by exposure to a long 
daily period of illumination, whether the light was wholly natural 
or partly artificial and of low intensity. 
Having shown that flowering was readily confined to either the 
upper or the lower portions of the stem by regulating the duration 
of the respective light periods, it was decided to study the somewhat 
more complex situation in which the central portion of the stem is 
exposed to a light period differing from that to which the upper and 
lower portions are subjected. Beginning July 11, the upper and 
lower portions of the stem of a Cosmos plant were allowed to receive 
only 10 hours of light daily while the central portion of the stem 
was exposed to the full period daylight. This was accomplished by 
using two light-proof cases, one arranged 12 inches above the other 
(pi. 3, i). Flower buds were visible on the lower portion of the 
K ant July 30 and on the upper portion July 26, while the first open 
ossom on the lower part appeared on August 21 and on the top 
part on August 11. No flower buds appeared in the central zone 
of the stem until the plant had been transferred to the greenhouse, 
except that at the node immediately below the base of the upper 
light-proof case a few buds appeared in late August. The flower 
buds at this node, which was scarcely an inch below the base of the 
dark compartment, were unable to develop and soon perished. 
Each portion of the stem responded to its particular light exposure 
in characteristic manner, without material interference from adjoin¬ 
ing sections responding in turn to a different light period (pi. 4, A). 
This experiment furnishes additional evidence that the action of a 
short light period in inducing flowering does not extend upward to 
parts of the plant simultaneously exposed to a long light period; 
and, similarly, this action in inducing flowering extends downward 
in the stem only a very short distance at most when the lower por¬ 
tion is exposed to a long light period. 
In another test begun July 13, the treatments just described were 
reversed, the middle portion of the stem being subjected to a 10- 
hour day and the upper and lower portions being exposed to the full 
length of daylight (pi. 3, a). Flower buds were visible on Sep¬ 
tember 10, on the central portion exposed to a 10-hour day, and the 
first blossom opened October 18. No buds appeared on the upper 
and lower portions exposed to the full length of day till after the 
plant had been transferred to the greenhouse. Because of the low 
height of the light-proof chamber the illumination within when the 
sides were open was poor and undoubtedly inadequate, and it was 
probably for this reason that unfolding of the flower buds was 
