131 
Origin and Development of the Composites. 
ridges and is crowned by about ten bristles which have lateral 
projections like those of Coreopsis. This solitary American repre¬ 
sentative would seem, therefore, to belong to the Helianthese, if it 
is a Composite at all. 
The large genus Cypselites, Heer, includes some undoubted 
Composite. Saporta’s determinations may be accepted as being 
accurate in most cases, but those by Heer require critical revision. 
Saporta left the position of the gypsum beds to be decided by later 
workers. Zeiller (37) gives the period as lower Oligocene, while 
Ward (36) gives it as Eocene. C. bisulcatus , because of the paucity 
of the ridges, and C. ellipticus, because of the numerous ridges, 
although possibly Cynareae, come under suspicion,which is increased 
by the curious projections from the beak in both cases. C. 
Lessiugii also has very numerous ridges and a very peculiar group 
of rigid hairs. C. rostnitus might also be excluded from the present 
discussion because of the absence of the characteristic ridges and 
the peculiar irregularity in length of the hairs. Some of these 
exclusions may seem severe but only thoroughly authentic Com- 
positae can be admitted for the purpose of elucidating the past 
history of the family. The genera Hieracites and Hyoserites contain 
several interesting species, all of which may be accepted as well 
defined Compositae. 
One of the most striking features of the well authenticated 
fossil Compositas is the predominance of the simple setose type of 
pappus. Although the American tertiary has as yet yielded practi¬ 
cally no Compositie, and the European forms, even from the Lower 
Oligocene, cannot be accepted as the earliest types, the evidence, 
as far as it goes, supports the view that the setose type of pappus 
is the primitive form (cp. Chap. V). The plumose setae of 
Cypselites Regelii and C. Ungeri, the outer setae with flattened, 
dilated bases (paleaceo-setose) of Hyoserites Schultzii,oi C. gypsorum 
and those of C. costatus which are connate at the base, are all com¬ 
paratively primitive types according to the view expressed in 
Chapter V. 
In connection with the writer’s opinion that the small, almost 
flat receptacle is the primitive type (see Fig. 19 and Chap. VI, B), 
it is noteworthy that the earliest known receptacula ( Hieracites 
stellatus and H. nudatus) are small and almost flat. Concerning 
the latter species, which is interpreted as a receptacle with only 
the scar of the involucre showing and no bracts, but a densely 
packed mass of florets or achenes, Saporta (32b, p. 57) says 
“ L’empreinte est difficile kinterpreter.” The figure (PI. 20, Fig. 5a) 
which he gives, however, is quite clearly a typical Cichoriaceous cap- 
