STOUT & Boas: STATISTICAL STUDIES IN CICHORIUM 351 
opment which give a periodicity or a sort of polarity. The proc- 
esses assigned by such conceptions as that of Ludwig to rhythmic 
cell divisions which give specific differences for species as such may 
themselves undergo change, continuous or discontinuous, in the 
development of successive parts of a single individual. 
VIII. SPECIAL VIEWS REGARDING HEREDITY, DIFFERENTIA- 
TION, AND SYMMETRY, HELD BY PEARSON AND BY BATESON 
While Pearson's studies of numerical qualities in plants do not 
pertain to number of flowers in any of the composites, they are of 
special interest in the recognition that differentiation is a factor 
in partial variability seen among organs of the same kind. They 
also illustrate very well the difficulties of adequately determining 
the heredity of such a character as the number of stigmatic bands 
or seed chambers in fruits of poppies, of Nigella hispanica, and of 
Майа rotundifolia. 
Pearson's earlier report ('or) bears on the statistical and 
mathematical demonstration that “undifferentiated like organs” 
ог“ homotypes”’ on an individual are alike only to a certain degree. 
The degree of likeness between homotypes as measured by his 
methods of determining homotypic correlation, has on the average 
a mean value of 0.4—0.5, which is, he considers, quite identical 
with the general value for fraternal correlation. Thus he con- 
cludes that heredity is a phase of homotyposis and that the 
sources of variability are to be sought in the individual. The 
distinction between differentiated and undifferentiated like organs 
is not, Pearson recognizes, always easy to make. In general, the 
former class involves function, position on the individual, season 
of production, etc., and is statistically discoverable by testing the 
frequency distribution for heterogeneity. Їп contrast to this, 
Pearson distinguishes variability of “undifferentiated like organs" 
as due to “that combination of small causes, inherent and en- 
vironmental, which leads to what is familiar in both theory and 
observation as a homogeneous chance distribution" (p. 287). 
We may note that when such differentiations as exist in the 
poppy and in Nigella are thus treated as pure chance variations 
the statistical treatment may give a high or low value for homo- 
typosis. The existence of differentiation is not necessarily re- 
