218 
James Small. 
as primitive.” Although he does not say which type is to be looked 
upon as primitive, the derivative position of the paleaceous type 
agrees with the views expressed above and in Chap. V on the 
primitiveness of the setose type. 
The Coreopsidinae are very closely allied to the Verbesininae 
and the primitive genus is probably Coreopsis , although the oligotypic 
genera Guizotia and Microlecane are even more closely allied to the 
parent group. As the centre of origin for all the sub-tribes of the 
Helianthese, except the Petrobiinae, is clearly the Mexican region, 
and as the two above-mentioned genera are African, the affinity is 
evidently another case of convergent evolution. The intimate con¬ 
nection between Bidens and Coreopsis indicates that these genera 
have been contemporaneous almost from the origin of the sub-tribe. 
The Galinsoginae are closely connected with Spilanthes through 
Calea, the chief genus of the sub-tribe, which has type IV style, 
type 4 or type 5 stamens, and differs chiefly in the numerous scales 
of the pappus, which are, however, sometimes as few as four, and 
narrow, rigid, acuminate, approaching the setiform tvpe of 
Spilanthes. 
The Lagasceinae and Petrobiinae are two small, somewhat 
anomalous groups, which in the absence of a detailed investigation 
can he taken as coming from the Verbesininae as in Fig. 7. 
The Madiinae, with the Layia-Madia group indicated as the 
primitive plexus, are more closely allied to the Galinsoginae than is 
suggested in Fig. 7. 
The Zinniinae, especially Zinnia , is so closely connected with 
the lower Verbesininae that it is impossible to give a precise source 
for the sub-tribe without more detailed investigation. 
The Ambrosiinae are more closely connected with the Melam- 
podiinae than with the Verbesininae as suggested in Pig. 7. The 
affinity between Iva (Ambrosiinae) and Parthenice (Melampodiinae) 
is so close that there can he no doubt of the systematic position of 
the Ambrosiinae in the Heliantheae and also very little doubt, if any, 
of the origin of the sub-tribe from the Melampodiinae via Partheniuvi , 
Parthenice , Cyclachccna (= Sect. I of Iva Hoffmann) and Iva. 
The affinities of the Verbesininae, Melampodiinae and Milleriinae 
are quite clear. Bentham(I,7, p. 434) points out that the Melam¬ 
podiinae are “ intermediate between Milleriese and the great mass of 
the Helianthoideae (Verbesineae), differing from the former in the 
completely paleaceous receptacle, from the latter in the constant 
sterility and undivided styles of the disk-florets.” 
