Origin and Development of the Compositei. 209 
margins of the involucral bracts. The epaleaceous recep¬ 
tacle and geographical distribution indicate the Chrysanthemidinae 
as the primitive sub-tribe, while these characters and those of the 
florets indicate Chrysanthemum as the primitive genus with 
Matricaria as more or less contemporaneous in origin. If C. 
segetum is carefully compared with a radiate Senecio with the 
calyculus well developed only the two chief tribal characters and 
general hahit are found to distinguish the two types, and C. segetum 
has a fairly extensive geographical area. 
Achillea is indicated by its floral characters and geographical 
distribution as the basal genus of the Anthemidinze, the chief 
difference being the paleaceous receptacle, which here, as in the 
Cichorieae, is considered to he an atavistic character. That it occurs 
in both cases in genera which are advanced in other characters 
isconfirmation of the reversionary nature of this type of receptacular 
appendage. 
The place of origin of the tribe is shown hy the present 
geographical distribution to be the Mediterranean region. The 
date of origin is indicated hy the same facts to be earlier than that 
of the Cichorieae. The presence of the Chrysanthemum-X\\\c leaves 
of Parthenites prisons in the Mediterranean region in the lower 
Oligocene, and the occurrence of the Anthemidean type of pappus 
and achene ( Hyoserites Scliultzii) in the Upper Miocene in another 
locality of the same region confirm, as far as is possible with our 
present knowledge, both time and place of origin. The Cypselitcs 
trisulcatus of the lower Oligocene has certain resemblances to the 
achenes of the Anthemideae and it is possible that further investiga¬ 
tion may show that those epappose achenes have in most cases been 
passed over on account of the absence of the pappus. 
Inulece. Throughout the analysis of the various characters of 
this tribe there has been some difficulty on account of the synthetic 
nature of the group. The diphyletic origin suggested by the 
filiform or bilabiate outer florets is confirmed by the structure of 
the involucre, the haustorial development of the antipodals, the 
secretory apparatus and the geographical distribution. 
If Tables VI and VI1 are examined in the light of these 
suggestions various points can be seen. In Table VI the 
Tarchonanthinae can be neglected as a small, special group. The 
other four sub-tribes with a filiform corolla (Plucheinae, Filagininae, 
Gnaphaliinae and Angianthinae) show a large proportion of type IV 
styles ; type XIII styles are also well represented in the “ filiform ” 
