354 SEMI-CENTENNIAL OF TORREY BOTANICAL CLUB 
tion between "chance variation" and “differentiation” in the 
treatment of homotyposis and heredity. He insists that '' meris- 
tic” variations are discontinuous and doubts that "there is a true 
material distinction between variation and differentiation as 
applied to parts of the same organism." Bateson further objects 
to a comparison of ‘‘undifferentiated like organs’’ with the cor- 
relation between brothers which may be differentiated as individ- 
uals. He evidently views the partial meristic differences of 
organs of the same kind actually in evidence in such plants as 
Nigella, Cichorium, etc., as a differentiation of the same rank as 
differentiation between individuals as such ('o3, p. 23). Bateson 
emphasizes the aspects of symmetry, advocates an extreme view 
that tissues and organs arise somatically by ‘‘differentiant or 
segregating divisions” in much the same sense as Weismann pos- 
tulated, and he thus questions the adequacy of the term “сһапсе 
variations." He is perhaps strongly influenced by his earlier 
studies of meristic variations in animals in which differentiation 
and symmetry are in marked evidence, and by the views of segre- 
gation of hereditary units representing characters which may give 
differentiation between individuals of the same hybrid origin. 
IX. EVIDENCE OF HEREDITARY VARIATIONS 
There appears to be no report of researches directed to the 
study of selection and heredity involving only total flower number 
per head in any of the Compositae. There are, of course, many 
species in cultivation from which double-flowered varieties have 
been developed, the history of which does not involve statistical 
studies of total flower number. Moreover, the development of 
so-called double-flowered composites does not necessarily involve 
increase or decrease of total flowers per head, but a change of such 
flowers as tubular disk-flowers into strap-shaped or ligulate 
flowers more like the ray-flowers. 
The studies of de Vries ('or) оп Chrysanthemum segetum are 
of interest in their bearing on selection, heredity, and evolution of 
flower number. He first isolated a race having a maximum of 13 
ray-flowers in the terminal heads. Then he isolated a race with 
21 ray-flowers in the terminal head. In this case, however, he 
considered it necessary to judge his plants not only by the ray- 
