Phloem in the Stems of Dicotyledons. II. 451 
capitulum character, and inflorescence-axis, peduncle, and vegetative 
stem would become dominated by this type of vascular structure. The 
presence of rudimentary, imperfect strands amongst the others or alone is 
explained on the assumption that certain factors arose, some of which are 
mentioned above, which induced a reduction amongst the constituents of 
the scattered system. 
It has also been suggested to the writer that while in Compositae the 
scattered system of the stem may have arisen by the above method, in 
other Natural Orders, where a quite similar scattered system obtains, a geo- 
phytic habit may have given rise to it in the lower region of the stem, and 
it would subsequently extend upward into the higher regions of the axis. 
In the same way the scattered disposition of the bundles in the petiole may 
have had an origin as stated above, while in other cases it may be due to 
the fact that there has been no mechanical necessity in that organ for the 
bundles to be arranged in a ring (owing, e. g., to the relative absence of 
bending-strains). Moreover, the scattered system may have arisen inde- 
pendently in both stem and petiole ; it may have already occurred in the 
petiole (if a fairly large organ) before the inflorescence became congested with 
its consequent appearance in the stem, and the scattered system in both 
organs would then have become mutually adjusted to produce the inter- 
dependent and continuous structure in the two organs such as we see, e. g., 
in the case of Cynara. This view involves the quasi-independent origin 
and development of stem and leaf. 
The ideas embodied in the above statements have, doubtless, on the 
face of them, a certain plausibility from the point of view of those who 
hold that stem and leaf have an independent and coeval origin (e. g. from an 
aboriginal thalloid ancestor). It is a question, between the hypothesis set 
forth above and that of the present writer, as to which is the most reason- 
able and likely to be true after the widest possible survey of all the available 
facts. For definite proof, either one way or the other, is not forthcoming. 
There can be no doubt that the inquiry into the real nature and origin of 
the vascular system of the Compositae is a very important one in view of 
the attention which this order attracts amongst botanists, the number and 
variety of forms it contains, and its wide geographical distribution. 
As already stated, the present writer holds the Compositae to be, as 
regards its vascular structure, a type of a very large number of Dicotyle- 
donous orders, and the conclusions drawn in this paper with regard to the 
nature and origin of its vascular system will apply, once and for all, not 
only to this order but to all others in which scattered bundles, medullary 
strands, or intraxylary phloem occur. For the writer regards these 
phenomena as essentially the same in every one of these orders, varying only 
according to the idiosyncrasy of the particular order in which they occur. 
He has investigated too many instances to harbour any longer a doubt on 
