452 Worsdell. — Origin and Meaning of Medullary 
this matter. Hence the reason why it is deemed necessary to give as full 
and complete a discussion as possible on the whole subject while treating of 
the Compositae. 
Now, in view of the belief just stated, and of various facts to be brought 
forward below, the writer regards the hypothesis of a dual origin for the 
scattered system 1 in the stems of Angiosperms as an extremely unlikely 
one. It seems much more probable that the vascular system of Angio- 
sperms has been founded and built up on a common ground-plan, and that, 
therefore, the scattered system has the same origin and meaning in all . 2 If 
not, then an adequate explanation should be forthcoming for the presence 
of medullary bundles in every Natural Order in which they occur. If their 
occurrence in Compositae is due to the congested type of inflorescence, 
then one might expect a similar explanation of their presence in those 
Umbelliferae, e. g. Eryngium , which also possess a congested type of 
inflorescence. As a matter of fact the scattered system obtains in the axis 
of the inflorescence of this genus. But the vegetative habit and external 
conformation of some species of this genus, e. g. E. Serra , is indistinguish- 
able from that of a Monocotyledon, and the scattered system of bundles in 
their stem and leaf almost entirely resembles that in- the stem and leaf of 
Monocotyledons. Hence the conclusion may fairly be drawn that the 
vascular structure in the stem and leaf of these plants has the same mean- 
ing and origin as it has in Monocotyledons, viz. a result of the geophytic 
habit, and is not an extension downward of that obtaining in the axis of the 
inflorescence, which, on the adaptive theory, is for the purpose of supplying 
the crowded florets. In this connexion it may be pointed out that in the 
axis of the inflorescence of E. amethystinum there is a system of medullary 
bundles which have nothing whatever to do with the supply of the florets, 
for they die out both above and below without forming connexions with the 
peripheral bundles, and therefore are clearly vestigial structures, having 
been retained in this region where bending-strains are absent, while, on the 
writer’s view, they have been purposively eliminated from the peduncle and 
vegetative axis of this species. 
In other, less Monocotyledon-like genera of Umbelliferae, e. g. 
Ferida , whose inflorescence is not congested like it is in Eryngium, the 
scattered system of the stem is equally well marked. It may be admitted 
that the complete absence of the scattered system in the stems of Dipsaceae 3 
and Proteaceae, both of which have congested inflorescences, may be due to 
the development of a hollow pith in the stem of some members of the 
former and to the woody; xerophytic habit of the stem in the latter. But 
1 In this part of the paper ‘scattered system’ and ‘medullary bundles’ stand for the same 
phenomenon. 
2 An exception must, perhaps, be made for the important group of the Amentiferae, which may 
have a distinct origin from other Dicotyledonous orders. 
8 It is present in the axis of the inflorescence of this order. 
