453 
Phloem in the Stems of Dicotyledons. II. 
this seems an inadequate excuse for their absence in Dipsaceae, which 
ought, one would think, to exhibit this feature if it is primarily due to the 
congested inflorescence, for in a stem of Morina elegans investigated by the 
writer there is a very wide solid pith, yet medullary strands are absent. 
Again, medullary strands are a very marked feature of the Melasto- 
maceae, occurring in numerous genera, yet there is no congested inflor- 
escence in this order, nor numerous floral parts, nor is there any other 
special character in the order which may account for their presence, unless 
it be the ring formed by the continuous bundles in the stem, ensuring 
adequate resistance to bending-strains, and therefore precluding the neces- 
sity of a hollow pith with the consequent elimination of medullary strands ; 
yet this cannot be the explanation, for Lasianthus strigosus , with a similar 
vascular ring, possesses no medullary strands. If the Melastomaceae have 
descended from arboreal or shrubby forms (which is the view generally 
held) the medullary strands would be, probably, a recently-acquired 
feature and adapted for some special function. But it is very difficult to 
see, on this view, for what function they have been acquired, why they 
occur in some forms and not in others, and why they are well-developed in 
some places and rudimentary in others. 
Again, to take at random two instances which are typical of others, 
why should the almost precisely similar intraxylary phloem-strands in the 
stems of Tragopogon (Compositae) and Nicotiana (Solanaceae) be supposed 
to have two quite different origins? For in Solanaceae a congested type of 
inflorescence does not occur ; hence some other cause for the presence of 
intraxylary phloem in this order must be found than that which is held, on 
the above view, to account for its presence in Compositae. Yet, in the 
present writer’s opinion, it is more than probable that a structure which is 
identical in both orders must have had the same origin in each. 
Nor, indeed, has the occurrence of phloem in the pith, such, e. g., as 
obtains in the Gamopetalae, ever yet been explained on purely adaptive 
grounds. Why phloem should occur in this (for it) wholly abnormal 
situation is a problem which requires solution, but which no anatomist has 
hitherto attempted to solve ! 
The sporadic occurrence of medullary bundles both in Compositae and 
other Natural Orders can with difficulty be accounted for on the theory 
which is opposed to that of the present writer. The case of Rudbeckia may 
be taken as an illustrative one. It might be assumed that the medullary 
bundles of R. maxima are retained owing to the greater diameter of the stem 
of that species as compared with that of most of the others, and that their 
presence in R. ( Echinacea ) purpurea is owing to the absence of a lacuna in 
the pith. But these differences in the stems of the various species appear to 
the writer to be insufficiently great or important, and therefore inadequate 
to account for the sporadic occurrence of the medullary bundles on the view 
h 1 
