196 Observation of the Origin of Varieties. 
WIG'S ray-curves 1 may also indicate the existence of a 
latent character, which, if it could be made active, might 
perhaps give rise to the double variety of the species. 
If we now examine the double varieties of the com- 
posites, we see that the structure of their inflorescences 
corresponds in every detail with that of Chrysanthemum 
segetnm plenum. The amount of doubling is always 
Fig- 35- Chrysanthemum coronarium. 
highly variable. The best cases reveal no tube florets 
as in the case of Chrysanthemum inodoniin in Fig. 34, 
p. 184. But if one looks between the ligulate florets 
small yellow tubes will be found in no inconsiderable 
quantity. This is true of Calendula officinalis and many 
other species. Such flowerheads are usually sterile, for 
1 May not the secondary maxima on the negative side similarly 
indicate the variety without ray-florets, the Var. discoidca (See 8, 
pp. 78-79 and Fig. 9, p. 80). 
