128 
James Small. 
number of short paleae ; as setiferous (Fig. 20, E) if the laciniae of 
the margin are long and narrow like the paleaceous setae of the 
pappus; as alveolate (Fig. 20, F) if the ridge is higher still with the 
margin entire or variously divided. 
In addition to these variations there is the paleaceous receptacle 
with membraneous scales, each of which subtends a floret. The 
latter point is frequently ignored and the term paleaceous applied 
to the setiferous receptacle but, as Cassini and Goebel maintain, 
there is an essential difference between the paleaceous setae, which 
surround the achene, and the paleae, which occur in the positions of 
bracts. 
Considering the origin of the capitulum from the racemose 
umbel, it is probable that the inner bracts were already very much 
reduced or entirely absent before the suppression of the pedicels. 
The appearance of scale bracts or other structures among the 
sessile florets would in that case be either reversion or a new 
development. The sporadic appearance of paleae in the position of 
bracts in genera or even species with all their relatives non- 
paleaceous has all the characteristics of atavistic variation. The 
series of variations from the foveolate to the alveolate and setiferous 
types shows a progressive development of the ridge between the 
depressions which clearly constitutes successive stages in the 
development of secondary structures. Their secondary nature 
is confirmed by their first appearance after the formation of the 
gyncecium in the development of the capitulum (see Martin, 11,47, 
p.355 and PI. XIX, Fig. 9b). 
The primitive receptacle would, therefore, be flat or nearly so 
and would show a foveolate structure, with reversion to a pre- 
Composite ancestor evident in the occasional appearance of paleae 
in the positions of bracts, and with the subsequent development of 
the setiferous and alveolate types indicated by the occasional 
appearance of the fimbrillate type. 
C. Analysis of the Receptacle Forms in the Composite. 
As pointed out at the beginning of this chapter the evidence 
from all the other parts of the capitulum is so strongly in favour fo 
the general scheme of evolution shown in Fig. 7 that the variation 
in the receptacle must be considered in relation to these other 
phyletic data. 
Senecionece. These data clearly indicate the Senecionese as the 
basal sub-tribe and Senecio as the basal genus. The condition of the 
receptacle in that genus is, therefore, very interesting. Bentham 
(1,8) describes it as flat or slightly convex, naked, foveolate or shortly 
