166 
James Small 
in different examples of the same species. This is only to be 
expected after Compton’s proofs (160) that in the Leguminosae 
the type of vascular anatomy in the seedling is correlated with 
the size of the latter. Hill and de Fraine (164), using examples 
from the Composite and other families, found that the area of the 
cross-section of the vascular bundles in the seedlings is correlated 
with the surface area of the cotyledons. Lee ( loc. cit.) found that 
in the Composite there are also variations which cannot be 
correlated with any external factors or characters. 
The cause of much of the variation in the anatomy of the 
seedling is thrown a step further back by Lubbock {op. cit., p. 102), 
who says “ The prevailing type of the cotyledons in the seedling 
stage is spathulate, a shape due to that of the seed, which in turn 
strictly conforms to the interior of the fruit.” The shape of the 
fruit clearly depends on the conditions under which it matures, for 
example, the fruits in the centre of the capitulum of most thistles 
are quite straight, but those around the outside are curved and 
have the pappus inserted obliquely. The narrow fruits of 
Scorzonera and Tragopogon mature on a flat receptacle, while 
enclosed by a long, rigid involucre. The result is the long, narrow 
cotyledon characteristic of these genera. In other genera where 
the fruits have more room to expand radially instead of longitudin¬ 
ally the short, broad type of cotyledon results. 
In the seedling, as in the embryo, Senecio shows the primitive 
type, and Lebard’s scheme (Table IV, Chap. I) must be reversed. 
This brings the genus Lactucci into the primitive group of the 
Cichorieae and the scheme, with some modification, is then in 
accordance with the views on the evolution of the Cichorieae given 
in Chap. X. With the Senecio or Lactuca type as primitive the 
variation in the other tribes, although in most cases considerable, 
is in accordance with the general lines of phylesis as deduced from 
other data. 
Pericarp, 
Various authors (61, p. 283; 169, 172, etc.) have investigated 
the detailed structure of the ovule, but Lavialle’s extensive work 
on the pericarp is the only one which yields interesting phyletic 
data. His conclusions are given in Pig. 3, Chap. 1, and are of 
interest as far as the Cynareae are concerned. The Centaureinae 
and Carduinae appear as a large basal group from which the 
Carlininae are given off, with Atmctylis as an intermediate genus 
and Xeranthemum connecting with the Mutisieae. The Echinop- 
sidinae are given as derived from the Carlininse (cp. Figs. 7 and 79). 
