Origin and Development of the Composites. 159 
would not readily he lost if the climber again developed the habit of 
an erect shrub or herb, especially if the phloem had become reduced 
in the interval. This would explain the relatively few cases, such 
as the Papaveraceae and the non-climbers in the above mentioned 
groups, in which latex occurs in comparatively low-growing herbs 
or shrubs. 
The present is not a suitable opportunity to develop this 
theory fully and we will proceed to apply it to the particular 
problem of the development of latex in the Compositae. Laticiferous 
vessels and the climbing habit are concomitant characters in the 
Siphocanipylus-Centropogon ancestors of Seuecio. In that genus 
and in all the Senecioneae w'hich have been examined there is a 
system of oleo-resin ducts throughout the root, rhizome and 
aerial stem (see Fig. 77). The explanation of the absence of 
food materials from the canals in this case is to be found in the 
facts given in Chap. XI, C-D. It was shown there that the 
dominant feature of the evolution of Senecio from the Siphocampylus 
group must have been the dwarfing of the plant in an Andine 
habitat and the aggregation of all the aerial parts. This is the 
very opposite of the phenomena which, on the above theory of 
latex, lead to the development of a laticiferous system. In these 
circumstances it is not surprising that the duality of function 
ceases and the laticiferous vessels degenerate into oleo-resin ducts 
with only an excretory function. For this reason an examination 
of Lysiponiiu from the latex point of view would be very interesting. 
Rhizocephalum (another Andine genus) and Apetahin, the other 
two genera which show considerable reduction in the gynoecium 
(cp. Chap. XI, C) have already been examined for latex with 
negative results (144 and 156). It is probable, therefore, that the 
change to oleo-resin canals had taken place during the dwarfing of 
the Andine plants before the origin of Senecio. 
We thus arrive at* a definite point of view with regard to the 
primitive condition of the secretory apparatus in the Compositae : 
i.e.y Senecio, being derived by a dwarfing process from the 
Lobelioideae in which a laticiferous system extends to root, rhizome 
and aerial stem, has that system modified by the loss of its nutritive 
function into a series of oleo-resin canals, which also extends to 
rhizome and aerial stem. 
Although many authors have contributed to our knowledge of 
the secretory apparatus of the Compositae (see Faivre, 125-6, Kny 
135, Triebel 149, Van Tieghem 151-3, Vuillemin 154-5, Leblois 136, 
Tr£cul 146-7 and Bibliography in Solereder 144) Col has given the 
most complete account (120) with numerous bibliographical refer- 
