Origin and Development of the Composite. 169 
have a bicellular pulvinus (Fig. II, Chap. V) which causes the hairs 
to diverge in moist conditions; the former are pauciseriate and 
have a pulvinus which causes the setae to diverge in dry conditions. 
This difference in the action of the pulvinus may be due to the 
position of the setae on the top of the pericarp and in any case is 
probably epharmonic. 
(6). Blake has pointed out (159, p. 6) the general concomitance 
of the presence of pappus and the presence of achenial hairs or 
the absence of pappus and the absence of achenial hairs. This 
“ pair of linked characters (pappus and pubescence of the achene).” 
is used by Blake {op. cit., p. 48) to separate two varieties of 
Viguiera flava, which are “sometimes growing together and not 
separable by any other character.’’ An extensive examination 
shows that the linkage of these two characters is very common 
throughout the Compositae in tribes, genera, species, varieties and 
forms. There are a few exceptions but the great predominance of 
the linkage is sufficient to render it almost, if not quite, certain 
that the two characters are inherited together. As they apparently 
behave as one character, there seems to he no real reason why they 
should be regarded as two distinct units in the genetic constitution 
of the plants. One systematist aptly summarised the point, when 
he said that “ one could not properly describe a pappose achene 
as glabrous.” 
Since there is no conceivable reason whythe character of a free 
calyx-limb should be linked with the character of pubescence of the 
achene, this new point is regarded as decisive. 
The teratological specimens of Treub (V, 66) and Worsdell, 
(IV, 96) are properly explained by Buchenau’s observations (see 
Chap. V, A) of a pappus inserted upon five green leaflets, which were 
developed in inverse proportion to the pappus. As the true sepal 
aborts the hairs upon it become larger until they are the only 
structures left. In this sense only can the pappus be regarded as 
a reduced calyx. This would also account for the occasional 
grouping of the setae in five more or less obscure bundles and the 
subsequent fusion of these bundles of setae to give five awns or 
paleae. 
Pubescence of the sepals is very frequently linked with 
pubescence of neighbouring parts such as the pericarp, the outer 
layers of which may be formed in the Compositae from the fused 
calyx tube. Another point in favour of this view is that where 
hairs occur on the corolla they may be organised into a pappus as 
in Leontopodinm (cp. Worsdell, IV, 95, p. 77 and Chap. V, A). 
Since all the facts adduced in support of the phyllome theory can 
