Origin and Development of the Composites. 77 
of structure and position are used to support the trichome theory. 
The view adopted in most text-books is that of Richard and Eichler 
that the pappus is a modified calyx limb, but it seems to the writer 
that Goebel’s argument (21, p, 317) concerning the nature of the 
receptacular setae is quite applicable in this case also. It runs as 
follows—“ Where now instead of single parts of a hypsophyll we 
see ‘ bristle-scales,’a qualitative change has taken place which may 
have begun with a transformation of the single parts of the 
hypsophyll, but I see no ground even then, if the hypsophyll no 
longer exists, for keeping its ghost hovering above, or rather below, 
these bristle-scales ; to assume that it still exists is an ‘idea,’ and 
this ‘ idea ’ is stuck somewhere in the axis and allows only the bristle- 
scales to appear. Such ‘ ideas,’ however, are to be found even in 
recent botanical literature.” 
This may be applied with more appropriateness on account of 
the evidence given below for the primitiveness.of the setose pappus 
within the family and the compound setose character of even the 
reduced number of paleaceous structures so frequently quoted as 
being very similar to a true calyx. 
B. Structure of the Pappus. 
When we examine the actual structure of the various types of 
pappus most of the difficulties connected with the trichome and 
phyllome theories disappear. We must, however, be content to 
regard the pappus as a structure more or less sui generis as advised 
by Richard. 
As with the styles and stamens we can reduce the in¬ 
numerable variations to a limited number of mean types which 
pass gradually into one another. These mean types will now be 
described briefly. The fundamental type is the scabrid seta, com¬ 
posed of uniseriate rows of cells fused together, with the obtuse 
terminal cell of each row free and projecting outwards as a lateral 
cilium for a distance which is less than the diameter of the seta. 
This is described as setose scabrid or setose denticulate (Fig. 17, A). 
That the structure is a fusion of uniseriate hairs such as are 
common on the cypsela is obvious and really requires no comment. 
There are various modifications of this basal type; if the 
ends of the constituent hairs do not project the seta is simple, as in 
Tussilago , and the upper part may be unicellular in cross-section as 
previously mentioned (61). This is described as simple setose , (Fig. 
17,B). The projecting lateral cilia may be acute and arranged in a 
