James Small. 
3 6 
The derivation of the Nassauviinae from the Senecioninae is 
confirmed by the bilabiate corolla which occurs in Brachyglottis 
(cp. Pig. 10, C-D), and by the predominance of yellow in the chief 
genera of the former sub-tribe. The spreading of the bilabiate condi¬ 
tion to the disc is quite homologous with the “ doubling ” which 
is of such frequent occurrence throughout the family, probably 
under the influence of an abundant food supply. The transition 
from the more or less homogeneous capitulum of the Nassauviinae 
to the forms with only several outer rows of bilabiate florets is seen 
in the genus Mutisia, which also shows a transition to the higher 
colours. The other genera of the Onoseridinse appear both from 
the form and the colour to be a special development, Mutisia itself 
giving rise to the Gerberinse. With this change both form and 
colour confirm the evolution of the sub-tribes as given in Fig. 7. 
In the Astereae it is clear from the colour relations of the sub¬ 
tribes that the relative positions of the Homochrominae and 
Heterochrominae must be reversed, as suggested in Chap. Ill, 
Sect. E, the other sub-tribes remaining as they are in Fig. 7. As 
the degree of development of the pollen-presentation mechanism in 
the two sub-tribes is very similar this can be done quite conveniently 
without disturbing the relationships of the other sub-tribes. We 
then get the yellow Homochrominae giving the Heterochrominae 
with the higher colours appearing first in the rays and then 
spreading to the disc. The filiform corolla appears more or less 
sporadically and can be interpreted as the expression of a tendency 
to reduction which reaches its highest development in these same 
genera. 
The spreading of the higher colours to the disc is a step 
towards the development of the discoid, highly coloured capitula of 
the Eupatorieae, which show higher colours as a rule, yellow 
appearing seldom, and then only in the basal group of the tribe. 
The disappearance of the ray florets is a normal phenomenon 
throughout the family. 
There is a great similarity in both form and colour in the 
Heliantheae and Senecioneae, but the higher colours are more 
frequent in the former tribe, thus confirming its position above the 
Senecionineae. The relative positions of the sub-tribes in the 
Heliantheae are neither confirmed nor refuted on account of the 
great variation shown in both form and colour. 
The derivation of the Helenieae from the Heliantheas is con¬ 
firmed by the similarity in form and colour, but the complete 
