449 
Phloem in the Stems of Dicotyledons. II. 
6. Within the same species, some individuals possess medullary strands 
in the stem, while other individuals are devoid of them, e. g. Crepis biennis , 
C. virens , Picridinm tingitanum } 
y. Where the medullary system of bundles is completely or almost 
completely absent in the stem, it is frequently conspicuously present in the 
leaf, e. g. Arctium Lappa , Crepis taraxacifolia , Senecio clivorum. 
8. There are plenty of genera in the order which do not exhibit, either in 
stem or leaf, the remotest trace of medullary strands ; but in almost all 
the features referred to under (i) and ( 2 ) are present, as they are also in the 
leaves of Rudbeckia, Tragopogon , and Scolymus. The vascular system of 
such genera is, on the writer’s view, the most advanced and has evolved out 
of that in which a medullary system was present. 
The course of the medullary system of strands through the main stem 
and the peduncle might have been described in the reverse direction, viz. 
from above downwards. If that had been done it would have been clearly 
seen that, as Kruch already pointed out, some of the medullary strands 
arise from the flowers , some from the vascular ring of the lateral branches 
of the stem, others from the vascular ring of the main stem. Great numbers 
arise de novo in the pith-tissue and may either pass into the vascular ring 
at a lower level or may form part of a system which ends blindly, i. e. dies 
out in situ below. Some individual strands are seen to leave the vascular 
ring and almost immediately rejoin it at a very slightly lower level. It 
would also be noticed that, on approaching the extreme base of the stem, 
a great reduction in the number of the medullary strands occurs either by 
means of extinction in situ , as mentioned above, or by anastomosis and 
copious fusion until a very few large amphivasal bundles, with well-developed 
xylem, occur at that level. Each of these bundles is the exact morpho- 
logical equivalent of the strand of internal phloem situated immediately on 
the inner, side of each bundle of the ring at a higher level in the stem. 
These bundles at a lower level either pass into the vascular ring, or, after 
losing their xylem, die out in situ. 
The medullary bundles are certainly not leaf-traces. The origin of 
some of them from flowers and branches has been stated above. Others 
derive from the vascular ring of the stem, which has been built up by leaf- 
traces ; such medullary bundles may be, therefore, regarded as leaf-traces 
of very indirect origin. The writer could find no evidence for the view set 
forth, e. g. by Weiss, that a particular medullary strand may be a leaf-trace 
which has entered the vascular ring from a leaf at the node above, pre- 
serving its individuality through the intervening internode till it enters the 
pith at the node below. 
The medullary strands of the stem are not continuations of those of the 
1 In P. tingitanum Kruch reports having found medullary strands, while the present writer 
could find none in the specimens he examined. 
