Origin and Development of the Compositce. 165 
the affinities which have been considered in previous chapters can 
be traced. 
As shown above, however, much more evidence can be adduced 
in favour of the phylesis as expressed in Fig. 7, Chap. II, and 
modified in subsequent chapters. The facts concerning the secretory 
apparatus are not only in agreement with the general lines of 
evolution but confirm a number of detailed affinities in a very 
interesting fashion. As this conclusion is arrived at by the use of 
the theory of latex outlined above it serves as a confirmation of the 
usefulness of such a theory, if not also of its actual truth. 
In spite of the extensive work by Col and others on latex in 
the Compositse, much remains to be done as our knowledge is still 
incomplete. This fact is well illustrated by the recent description 
(116) of latex sacs in a species of Parthenium (Melampodiinae), one 
of the advanced Heliantheas, in which latex is present in sufficient 
quantity to be of possible commercial value as a source of Guayule 
rubber. Another source of rubber is Hymenoxys doribundci 
(Heleniinas) but the quality is said to be inferior (see 119a). 
D. Some Isolated Data. 
There remain to be considered some isolated data referring 
to morphological and physiological characters which are of minor 
importance or which do not extend to a representative number of 
genera and tribes. 
Seedling Structure. 
The work of Dufour (1, 26) and Lebard (I, 53), which was 
mentioned in Chapter I, is now of more interest. Lebard derived 
the Cichorieae from the “ Tubuliflores ” through the genera with 
long, narrow cotyledons, the Leptocotylees of Dufour. The 
examination of an extensive and representative collection of 
seedlings grown for the purpose shows that the broad cotyledon 
is the predominant type throughout the Tubuliflorae ; it may be 
long or short but only in a minority of genera do we find the long 
cotyledon also linear and in still fewer is the short cotyledon at all 
narrow. Lubbock (167) considered that “One of the simplest 
types of embryo is that seen in the species of Senecio ,"where the 
cotyledons are short and oblong-obvate. 
Lee’s work (1, 54) on the seedling anatomy of the Composite 
led him to the conclusion that such data are of no use in questions 
of affinity. There is a marked divergence of opinion on the primitive 
type of root, diarch or tetrarch (cp. Thomas, 170). Only one type 
of transition with variations occurred in the fifty species of 
Compositae examined ; these were all diarch or tetrarch and showed 
variations between these extremes in nearly related species or 
