450 Worsdell. — Origin and Meaning of Medullary 
leaf, which has its own independent system. On the other hand, they are 
sometimes, as in the cases of Scorzonera and Helminthia , direct continua- 
tions of those of the branches. 
Discussion and Conclusions. 
The position assumed in this paper is that the Compositae are primi- 
tively and aboriginally herbaceous plants in habit and consistence, a view 
which is borne out by the fact that shrubby and arborescent forms are rare 
in the order. It is in this way that the writer accounts for the vascular 
structure of the stem and leaf throughout the order. The very irregular 
alinement of the bundles composing the vascular ring of the stem , 1 the 
presence in so many forms of medullary bundles or a scattered disposition 
of the ring-bundles both in stem and leaf, are brought forward as evidence 
in favour of the above view. 
It seems to the writer a matter of considerable importance to try and 
determine what was the primitive type of vascular structure of the vegetative 
organs of the Compositae, and to endeavour to settle the question as to 
whether the medullary or scattered system of bundles, occurring either in 
stem or leaf, is a comparatively recent, adaptive structure or whether, on 
the contrary, it is an ancient, vestigial one. For the structure, viewed as 
a whole, must have one or the other origin. 
What are the arguments in favour of what may be termed the ‘ adap- 
tive theory ’ ? 
It has been suggested to the writer that the presence of medullary or 
scattered bundles in the stem may be the result of the acquirement of 
a congested ' ype of inflorescence in this order known as the ‘ capitulum ’. 
This particular type of vascular system would result, in the axis of the 
inflorescence, from the excessive shortening of the internodes in that region, 
combined, possibly, with a relative increase in the diameter of the axis. 
(As a matter of fact, it regularly occurs in the axis of the congested 
inflorescences of Compositae, Umbelliferae, and Dipsaceae.) The scattered 
system there established might then extend, in more or less accentuated 
form, throughout the peduncle and the vegetative stem. If, however, inhibit- 
ing factors prevailed, such as the reduction in diameter of these organs, the 
necessity for resisting bending-strains in an elongated axis, or the formation 
of a hollow pith (central lacuna), medullary bundles would either not be 
formed at all, or in very reduced numbers, or rarely. Further, if the 
scattered system occurred in the vegetative stem from the cause just men- 
tioned this might induce, by correlation, a similar structure in the petiole of 
the leaf. 
The scattered system of bundles would have been first laid down when 
the original loose spicate or racemose inflorescence first began to acquire the 
1 The reader must be again reminded that the structure is described from the transverse section. 
