Origin and Development of the Composites. 273 
occurs in the tactile hairs when they are touched, thus starting the 
changes which result in the movement. 
It is hoped that the necessary apparatus will be devised 
presently to prove the similarity, if it exists, in the stimulation of 
the root and of the stamens, and it may be possible also to test a 
suggestion that in both cases the first result of the application of 
the stimulus is a change in the electrical condition of the limiting 
layer of the protoplasm, which change alters the permeability, and 
therefore the turgor, giving the elastic cell-walls an opportunity to 
contract in the case of the filaments and to expand at a relatively 
greater rate on the upper side in the case of the root. 
E. The Phylooenetic Significance of Irritability 
Considering all these facts in the light of the probable lines 
of evolution (found as a result of the examination of the morphology 
of the styles and stamens) shown in Fig. 7, Chap. II, the first 
point to be noticed is that in no case is a sub-tribe which shows 
irritability derived from one showing no irritability. This fact in 
itself is very good proof that irritability has followed the same 
lines of evolution as the morphology of the pollen-presentation 
mechanism. The second point is that the possible affinities of 
the Gnaphaliinae with the Gorteriinae and Gerberinag are rendered 
less probable by the occurrence both in the Gorteriinae and 
Arctotidinae of irritable styles of the same type, and by the occurrence 
in both the Nassauviinae and Gerberinae of an explosive mechanism 
of a peculiar type. These two pairs of sub-tribes seem to be more 
intimately connected than is suggested by the morphology of their 
styles and stamens. The development of the A type in several 
distantly related groups is not of great importance, as that type of 
movement is a logical development in the evolution of irritability. 
The Cichorieae are again isolated on account of the peculiar, 
slow movement of the anther tube, which movement, however, is 
also shown by Senecio Ligularia, var. speciosa. 
The Tussilagininae, Vernonieae, Eupatorieas, Ambrosiinae and 
Athrixiinae are all end groups derived from groups in which irrit¬ 
ability is present, so that the absence (as yet) of records of irritability 
for these groups does not in any way invalidate the previous 
arguments. As before the Othonninae cannot be distinguished 
from the Senecioninae, nor can the various tribes of the Cichoriese 
be separated from each other with any certainty. 
The Calenduleae show a marked development of the C type in 
Dimorphotheca, thus confirming the affinity with the Senecionime. 
The Arctotideae are again isolated on account of their irritable 
