Origin and Development of the Composite. 29 
rubber bands were used. Strong spring clips also were used to 
obtain pressure on two sides of the capitulum. 
The material used was Cotula coronopifolia (a discoid species) 
Ursinia speciosa, U. pulchra, Chrysanthemum Leucanthemum (rayed 
species), and Dimorphotheca pluvialis; this last species had given 
numerous capitula with bilabiate corollas in the first and second 
rows of disc florets. 
In no case did the treatment make any difference in the 
development of ray florets but in a few cases of C. Leucanthemum 
tied with twine the central florets of the disc were obviously 
crushed and failed to mature. The pressure in such a case arises 
from the expansion of the florets as it would in the case of the 
disc florets pressing the rays against the more or less rigid 
involucre. 
The influence of the food supply was also investigated in C. 
Leucanthemum and Dimorphotheca pluvialis. This was done by 
cutting a notch halfway through the stalk of the inflorescence, 
protecting the wound with vaseline and supporting the capitula. 
The results were again rather indefinite, but in the latter species 
the capitula so treated were rather poorly developed as a whole 
without much difference in the rays. In one case, however, there 
were five buds in different stages above the wound and, while the 
other peduncles of the same plant gave numerous cases with 
bilabiate disc florets, none of the capitula with deficient food 
material showed this modification, which involves a greater 
expenditure of material. 
These results suggest that the factor for radiateness is 
probably a Mendelian one (see also Trow, 84) acting directly on the 
production of rays and not indirectly on the food supply or the 
amount of pressure developed by the disc florets. An abundance 
of food supply is suggested, however, as a contributary cause in 
the production of bilabiate disc florets and this explains, in part 
at least, the frequent production of “ double ” flowers by cultivated 
Composites. 1 
D. Variation in Colour of the Corolla. 
The phylogenetic value of colour as a character depends upon 
two chief points, our knowledge of the relation of colours and 
insects and our knowledge of the composition of the pigments. In 
1 De Vries (87, p. 507) records warm weather as a factor in the production 
of “ doubling ” in the cornflower. 
