I 20 
James Small. 
The only point now to be determined is whether the capitulum 
arose from a spike or from a racemose umbel. The smallness and 
flatness of the receptacle, the very reduced character of the 
receptacular bracts when they are present, the small size of the 
individual florets and the dominant tendency to zygomorphy of 
the outer, florets are all evidence in favour of the origin of the 
capitulum by the suppression of the pedicels in a racemose umbel. 
The only evidence in favour of the origin from a spike is the 
occurrence of an elongated receptacle together with receptacular 
bracts in a few of the Heliantheae, such as Rudbeckin , but these are 
quite advanced in the characters of the style, stamens, corolla and 
pappus, so that they can scarcely be considered primitive types. 
This derivation would also leave unexplained the small size of the 
florets, the dominant tendency to zygomorphy in the outer florets, 
the reduced condition and frequent absence of the receptacular 
bracts and, more important still, the dominance of the small, flat 
or very slightly convex receptacle. All these characters and 
tendencies are more readily intelligible when the racemose umbel 
is taken as the ancestral inflorescence. 
The view adopted here of the origin and development of the 
involucre in the Compositae is briefly as follows: The ancestral 
racemose umbel by abortion of the pedicels gave a capitulum with 
the bracts of the outer florets forming a uniseriate pericline, while 
aggregation of the cauline leaves immediately below the inflorescence 
gave the calyculus. The primitive involucre has, therefore, an 
uniseriate pericline with a calyculus composed of a few scattered 
members. Progressive sterilisation of the floral bracts has given 
the pluriseriate pericline, while progressive aggregation of the 
cauline leaves has given a denser, pluriseriate calyculus. 
It must be observed, however, that in some cases the caly¬ 
culus passes gradually into the pericline, but that is a question of 
phyllotaxis and is considered in Chapter VIII. 
C. Analysis of the Involucre in Composite. 
Taking the uniseriate pericline and the slightly developed 
calyculus as primitive we can now analyse the distribution of the 
various types of involucre in the family. 
Scnecionece. The genus Senecio again shows the primitive type 
with a number of variations. The pericline is uniseriate with the 
bracts free or more or less connate at the base, or there may be a 
smaller number of “ bracts ” in a second row. These outer 
“ bracts ” are probably a development of the smaller leaves of the 
