168 
James Small. 
senecine, are recorded for Senecio vulgaris (161). The intimate 
relationship of Senecio and Ligularia is emphasised by the 
occurrence of senecioic acid in Ligularia tussilaginea (157), while 
the less close relationship of Arnica and Tussilago is confirmed by 
the occurrence in the latter of faradiol, a bivalent dextrophytosterol 
closely related to arnidiol, a constituent of Arnica (165-166). 
Many of the constituents of the latex, oleo-resins and gum-resins 
have a commercial value, but the most interesting case is 
Stevia Rebaudiana, the dried, powdered leaf of which is 40-50 times 
sweeter than any other natural product and can be used in the 
crude condition as a non-toxic substitute for saccharin. The sweet 
constituent is a glucoside, estevin or eupatorin, and the sweetening 
power of the pure substance is 150-180 times that of cane-sugar 
(see 171 for review of facts and literature)* 
Much more is known of the chemical constituents of the 
Compositae but much more still remains to be added before the 
facts can be used in a rational manner, either in the investigation 
of phylesis or in the utilisation of the plant products for economic 
purposes. 
Pappus. 
Since Chapter V was written some adverse criticism has been 
made of the views there expressed on the trichome nature of the 
pappus. Further evidence has arrived also,which supports the trichome 
theory in a very interesting way. As no decided opinion on the 
fundamental nature of the pappus was expressed in the previous 
account, the issue will now be made clear. The present writer holds 
that the pappus is a trichome structure or an emergence and not a 
divided calyx, for the following reasons:— 
(1) . The structure of the mature pappus, even in the 
paleaceous forms, is that of a series of hairs which have become 
fused throughout all or part of their length, either side by side to 
give a scale or in a mass to give a seta or awn. 
(2) . The development of the members of the pappus is either 
that of a typical trichome or that of an emergence, such as the 
surface spine of the thistle-leaf, which is comparable with some of 
those anomalous cases in which vascular bundles have been found 
in the pappus. 
(3) . The primitiveness of the scabrid seta is in conformity 
with the evolution of the family as deduced from other data. 
(4) . The predominant type of pappus in the fossil forms is the 
setose type. No fossil foliose pappus is known. 
(5) . The similarity of the setae to the achenial hairs is very 
marked. In the primitive genus the latter are aheady.biseriate and 
