Phloem in the Stems of Dicotyledons. II. 455 
forrrf, the adaxial portion of the vascular ring being frequently also in 
a correspondingly rudimentary state (Fig. it). 
The genus Lactuca , as regards the occurrence of medullary strands in 
stem and leaf, exhibits interesting variations. L. Plumieri , e. g., has almost 
lost them in the stem, whereas they are abundant in the leaf, although in 
a rudimentary form. 
In L. mctcrophylla they are completely absent in the stem, and only 
the slightest remnants of them can be found in the leaf, a fact which is 
obviously due to the presence of a very large lacuna in the pith. Where 
the lacuna is quite small or absent, medullary strands are usually present, 
as in L. Plumieri and L. hastata ; in the last-mentioned species medullary 
strands are entirely absent from the stem. In L. Scariola medullary 
strands are exceedingly well developed in the stem ; in the leaf they are 
quite absent. 
This last fact shows that, although the leaf is a more conservative 
organ than the stem, there are cases in which its vascular system has 
become more reduced than that of the stem. The medullary strands in the 
stem of L. virosa , L. saligna , L . Scariola , and L. sativa have been retained 
in order to subserve some useful function ; they are not merely vestigial 
but, in all probability, also adaptive structures . 1 
In the leaf no necessity for their retention exists, and at the same 
time this organ, as compared with that of L. Plumieri and L . alpina , has 
become reduced in size. The same phenomenon recurs in the case of 
Rudbeckia. In some species medullary strands occur in the stem, in other 
species they are absent. But in the leaf of all species except R. {Echinacea) 
purpurea they are completely absent, the entire vascular system having 
been reduced to a comparatively short arc of bundles, a fact which is 
probably due to the petiole having been relatively reduced in size . 2 
A striking example again is afforded by Tragopogon . The stem of all 
species has very numerous medullary strands. The leaf has become much 
reduced and almost grass-like in its organization, in accordance with which 
fact the medullary strands, in the case of T . pratensis , have become quite 
extinct in the leaf, while in the case of the other species investigated they 
are small and insignificant as compared with those of the stem, dying out 
in the leaf-base. 
The theory, therefore, is that in Compositae the vascular structure of 
the leaf, wherever it is large and well developed, constitutes a reliable index 
to the fact of the former universal presence of medullary strands in the 
stem. But it must be taken in conjunction with the fact of the prevalently 
imperfect, rudimentary character of the medullary system of the stem 
(wherever this occurs), and with the phyton-theory as a basis (viz. that 
1 This must be true in all cases where the medullary strands consist of functional tissues. 
2 The lamina may be of quite large size. 
L 1 % 
