Origin and Development of the Composite. 29 
distribution of Adenostemma viscosum (Piqueriinae) is clearly the 
result of its very effective means of dispersal (IV, 98.) 
Helinnthece. The wide areas of a number of genera in the 
Verbesininae (Fig. 39) and the slightly less wide distribution of 
the Coreopsidinae confirm the basal positions of those two groups. 
The Mexican centres and distribution of the Madiinae, Zinniinae 
and Lagasceinae, and the Mexican-West Indian development of 
the Milleriinae confirm the positions of these sub-tribes. The 
Chilian centre of the Petrobiinae co-incides with a considerable 
development of the Verbesininae. The position of Petrobium on 
St. Helena can only be accounted for by a polyphyletic origin of 
this sub-tribe : the probability of polyphylesis will be considered 
in the next chapter. The wide distribution in America of the 
Galinsoginae and their diffuse development from the U.S.A. to the 
Andes confirms the direct origin of this group at a comparatively 
early date from the Verbesininae. 
Heleniece. Except for their almost complete absence from the 
U.S.A. the Tagetinae show a remarkable similarity in their distribu¬ 
tion to that of the Galinsoginae, so that the origin of the former 
from the latter sub-tribe is more or less confirmed, but the facts of 
distribution would also support an origin of the Tagetinae from the 
Senecioninae as suggested in Chapter V, D. The Mexican centres 
and areas occupied by the other sub-tribes are in accordance with 
their relationships, as suggested in Fig. 7 and modified in Chapter IV, 
D. The extension of fawned and other details of the distribution of 
the Jaumeinae confirms the suggestion (Chapter IV, D) that this sub¬ 
tribe arose not later than the Baeriinae. The independent origin 
of the Heleniinae suggested in Fig. 7 and confirmed in Chapter III, E, 
is in accordance with its distribution, which is more like that of the 
Baeriinae than that of the Tagetinae. 
Conclusions. It is clear, therefore, that (except for the 
derivation of the Conyzinae from the Homochrominae and the 
reversal of the positions of the sections of the Gnaphaliinae) the 
views on the origin and development of the tribes of the Com- 
positae which have been given in Fig. 7 (Chapter II) and modified 
slightly in subsequent chapters receive complete confirmation in 
considerable detail from the study of the geographical distribution 
of the family. 
The value of the graphic method of using maps for each genus 
or tribe, instead of the statistical method used by Bentham, is 
clearly proved. 
