Stout & Boas: STATISTICAL STUDIES IN CICHORIUM 347 
Detailed studies by Burkill ('95) show that the number of 
stamens in flowers of Stellaria media increases during the first two 
weeks of bloom, then there is a gradual decrease to the end of the 
flowering season. Reinóhl ('o3) working with the same plant 
showed further that the plants grown in poor soil produced flowers 
with a lower number of stamens. Tammes’s (705) studies of the 
effect of good and poor soil conditions on various characters show 
lower values for many of the characters of poorly nourished plants. 
Love’s researches (’11) are among the most recent that deal with 
the effect of nutrition on the mean of such characters as height, 
number of peas per pod, weight of seeds, etc., in such plants as 
peas, buckwheat, and corn. His results agree with those men- 
tioned above, that is, he finds an increase in the mean and in the 
variability, as a result of better nutrition. 
VI. STUDIES OF INTERANNUAL VARIABILITY 
In the consideration of interannual variability, at least two 
distinct aspects are to be recognized. First, the season of growth 
may differ in a way that affects the plant, and, second, in the case 
of perennial plants the age of the plant may be a factor in vari- 
ability just as the period of development may influence the partial 
variability seen in a single year of growth. These two factors 
have rarely been distinguished and little attention has been directed 
to the factor of age, for flower number studies have been chiefly 
indiscriminate for the population. 
Haacke (706), it seems, is the only investigator who has 
attributed variation in the number of ray-flowers of composites 
to the age of the plant. He suggests that older plants probably 
have more ray-flowers per head. His observations on Chrysan- 
themum Leucanthemum and Anthemis arvensis were not conclusive, 
however, for he did not know the age of the plants studied. 
Yule (7о2) gives the results of counts made in three successive 
years of the number of sepals in a population of Anemone nemorosa 
from one habitat. He zets differences in the means for different 
years and calls attention to the importance of observing ''local 
races" for several years before one can determine the character- 
istics of a species. Shull ('o4) made a similar study for a popu- 
lation of Aster prenanthoides and obtained higher mean values for 
