454 Worsdell . — Origin and Meaning of Medullary 
that these are to be regarded as a comparatively new and adaptive feature. 
It is, on this view, still more difficult to account for their presence in the 
stems of some species of Lactuca 1 and their complete absence in other 
species, and the writer is unable to discover any adaptive features which 
should explain this remarkable fact. 
Again, the writer cannot agree that the scattered disposition of the 
bundles in the petiole (where it occurs) can have two distinct modes of 
origin ; here also, there must be a single underlying cause for the pheno- 
menon not only in the Compositae, but in all other Dicotyledons which 
exhibit it. A survey of all the facts strongly points to this conclusion. 
The hypothesis, as stated above, that the scattered system may have 
had a quite independent origin in stem and leaf is, on the view that these 
organs are quasi-independent the one of the other, a quite possible one. 
But, having regard to all the facts in all Natural Orders concerned, it is 
a very inadequate hypothesis, especially in view of the intimate connexion 
and relationship between stem and leaf. A single fact which is detrimental 
to this view may be cited, viz. the continuation of typical leaf-structure, in 
so many cases, into the cortex of the stem. 
The theory held by the writer is that the leaf, in its anatomical organi- 
zation, is identical with that of the stem ; that it has an epidermis, cortex, 
vascular system, and pith which are morphologically identical, and aborigi- 
nally one and continuous, with the similar tissues of the stem. The morpho- 
logical and anatomical history of the two organs are inseparably bound up 
together. Further, the leaf is a more conservative organ than the stem. To 
cite a single physiological fact : the leaf has not been obliged, to the same 
extent as is the case with the stem, to adapt its structure for the resistance 
of bending-strains. This being so, and if the medullary system of strands 
of the stem represents, as seems obvious for the reasons set forth above, an 
ancestral character, then we should expect this system to be more frequently 
present in the leaf than in the stem or, if it occurs in the stem, that it 
would often occur in a better developed condition in the leaf. This is, in 
fact, exactly what we find. 
In the stem of Senecio medullary bundles are practically extinct — only 
extremely rudimentary traces of them were observed in A. Petasites. Yet 
in the petiole, for example, of N. clivorum , medullary bundles represent an 
extremely well-developed system. A similar state of things exists in the 
case of Inula Helenium . 
In the genus Crepis medullary bundles are either absent or very rare in 
the stem ; whereas in the petioles of all species examined by the writer 
they are present in considerable numbers, although in a very rudimentary 
1 They probably have some use, or they would not have been preserved ; but it is much more 
easy to account for their distribution in this genus on the vestigial theory than on the theory that 
these strands are recently-acquired adaptive structures. 
