Origin and Development of the Composite. 163 
position of the Anthemideae, Heliantheae and Helenieae is in 
accordance with Pig. 7, except that a close affinity has again to be 
noted between the Senecioneae and Helenieae (cp. Chaps. V, D and 
X, D). The somewhat abrupt disappearance of the canals from 
the stem in the Calenduleae and Arctotideae ( Ursinia and Arctotis) 
is not surprising when the variation in a single tribe (as in the 
Cynareae, Mutisieae, etc.) is observed, 
In the Vernonieae two distinct types occur; the appearance 
of laticiferous sacs in Vernonia may be correlated with the climbing 
habit of many of the species and a thorough examination of the 
climbing species of Senecio would be of interest in this connection. 
The derivative position of the Lychnophorinae is confirmed by the 
total absence of secretory canals from the stem. 
The distribution of the canals in the Inuleae raises again the 
question of the monophyletic origin of that tribe. The “ Gnaphaliees 
of De Candolle ” includes the first seven sub-tribes of the Inuleae 
of Bentham. The facts indicate a diphyletic origin, as suggested in 
Chaps. IV, P and VI, C. The variation in Inula would account 
for the distribution of the secretory apparatus in the Athrixiinae 
and Relhaniinae. 
One of the most interesting points is that the secretory 
apparatus is equally developed in the Senecioneae, Inula Helenium 
Buphthalmum and the Centaureiirae. The only difference is that in 
the first and last groups the canals are opposite the vascular 
bundles, while in the other two they are placed laterally or 
between the bundles. The origin of the Cynareae from the 
Buphthalminae thus receives comfirmation. In Cynara and 
Saussurea (Carduinae) the development of the canals is the same as 
in the Centaureinae, so that the position of the Echinopsidinae as 
derived from the Carduinae (Figs. 7 and 77) is confirmed. The 
Carduinae show a gradual development of laticiferous sacs 
possibly connected with the ecological conditions which have 
played so conspicuous a part in the development of the Cynareae 
(cp. Col’s conclusion from his experiments that the environment 
determines the appearance of the secretory organs in some 
Compositae but notin others (120d, p. 164). The change in the case 
of the Cynareae would be rendered easy by the ancestral laticiferous 
characters, which, although suppressed, do not appear to have 
been completely lost. The Carlininae show a wide range from 
Carlina acanthifolia with anastomosing laticiferous vessels to C. 
caulescens, Xerantheninin and Cardopatium. This is in accordance 
with the derivation of this sub-tribe from the Carduin* (Fig. 7), 
