8 9 8 
Journal of Agricultural Research voi. xxm, No. h 
of tubers occurs with a light period which admits of only very limited 
stem growth. The absolute maximum production of tubers, however, 
is obtained with an intermediate light period, or, as actually occurs 
each year in higher latitudes, with a light period favorable to extensive 
stem growth followed by a period favorable to intensive tuber formation. 
In all cases thus far studied typical tuber formation takes place as a 
result of decrease of the daily light period below the optimum for stem 
elongation so that this type of tuberization is greatly favored by the 
shortening of the day in fall following the long summer days of high 
latitudes. Bulb formation, as illustrated by the onion, on the other 
hand, takes place as the result of increase of the daily light period above 
the optimum for stem growth so that formation of these structures is 
favored by the long days of summer which occur in high latitudes. 
The results of the experiments in tuberization seem to show that 
the remarkable regulatory action of change in the length of day on 
stem elongation and various associated phenomena cannot be due 
directly to change in the quantity of carbohydrate formed through 
photosynthesis. The content of plastic carbohydrate in the plant 
at any given time, of course, will depend on the relation between rate 
of formation through photosynthesis and the rate of consumption in 
promoting growth and other vital functions. Tuberization furnishes 
visible evidence that under the prevailing conditions production of 
total carbohydrate is in excess of current consumption except in so far 
as local tuberization may be the result of mere transfer of preformed 
storage material from other parts of the plant. There seems to be little 
doubt but, beginning with the optimal day length for stem growth, 
decrease in length of day in general involves decreased formation of 
carbohydrate. It appears, however, that progressive loss of power 
to grow as a result of decreasing day length advances more rapidly 
than would be required from the decreased formation of carbohydrate 
as a controlling factor. It is necessary, therefore, to conclude that 
the duration of the light period exercises a regulatory action on internal 
processes of the plant other than those which merely determine the 
total quantity of carbohydrate produced. It remains to be determined 
whether the additional action is to regulate the form of carbohydrate 
produced or to control other processes not so clearly connected with 
carbohydrate formation. In any event there is convincing evidence 
that the duration of the daily light period is an important, and often 
a controlling, factor in determining whether the plant may use the 
product of photosynthesis as rapidly as elaborated in promoting growth 
or, instead, must store the material, perhaps for its own subsequent use 
or for the nutrition of its sexually or vegetatively produced progeny. 
CHARACTER AND EXTENT OF BRANCHING 
One of the characteristic effects of altering the light period from 
optimum to suboptimum for apogeotropic response is to promote branch¬ 
ing. This increased tendency toward branching is frequently associated 
with flowering, a fact which is not surprising since both are promoted by 
a change in the same direction in the duration of the light period. 
Branching and flowering are not always associated, however, for not 
merely the direction but also the extent of the change in the light period 
must be taken into account. Apparently the most favorable light period 
for aerial branching lies between the apogeotropic optimum and that 
